From c9f5b25caec1b11990bf816a6a21513bd2fb76c7 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Mon, 30 Apr 2018 18:23:58 +0200
Subject: [PATCH 01/28] Add files via upload

---
 Work/CNN_XRAY.py                              | 227 ++++++++++++++++++
 Work/NN-Xray.py                               | 112 +++++++++
 Work/numpyrec/NN.py                           | 100 ++++++++
 Work/numpyrec/nprecord.py                     |  64 +++++
 Work/tfRecord/Tfrecord.py                     | 199 +++++++++++++++
 Work/tfRecord/__init__.py                     |   2 +
 .../__pycache__/Tfrecord.cpython-36.pyc       | Bin 0 -> 3599 bytes
 .../__pycache__/__init__.cpython-36.pyc       | Bin 0 -> 205 bytes
 .../decode_tfrecord.cpython-36.pyc            | Bin 0 -> 3051 bytes
 Work/tfRecord/decode_tfrecord.py              | 128 ++++++++++
 Work/utils_data/norma/norma.py                |   7 +
 11 files changed, 839 insertions(+)
 create mode 100644 Work/CNN_XRAY.py
 create mode 100644 Work/NN-Xray.py
 create mode 100644 Work/numpyrec/NN.py
 create mode 100644 Work/numpyrec/nprecord.py
 create mode 100644 Work/tfRecord/Tfrecord.py
 create mode 100644 Work/tfRecord/__init__.py
 create mode 100644 Work/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
 create mode 100644 Work/tfRecord/__pycache__/__init__.cpython-36.pyc
 create mode 100644 Work/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
 create mode 100644 Work/tfRecord/decode_tfrecord.py
 create mode 100644 Work/utils_data/norma/norma.py

diff --git a/Work/CNN_XRAY.py b/Work/CNN_XRAY.py
new file mode 100644
index 0000000..90da2f1
--- /dev/null
+++ b/Work/CNN_XRAY.py
@@ -0,0 +1,227 @@
+import tensorflow as tf
+import tfRecord as trec
+
+
+T_HEIGHT = 32
+T_WIDTH = 32
+T_channels = 3
+
+TRAIN_FILE = 'tftrain.tfrecords'
+TEST_FILE = 'tftest.tfrecords'
+
+learning_rate = 0.001
+
+num_steps = 1500
+
+batch_size = 8
+
+display_step = 10
+
+
+
+# Network Parameters
+
+num_input = T_HEIGHT * T_WIDTH * T_channels
+
+num_classes = 2 
+
+
+
+
+# tf Graph input
+
+X = tf.placeholder(tf.float32, [None, num_input])
+
+Y = tf.placeholder(tf.int64, [None, num_classes])
+
+
+
+
+
+
+
+# Create some wrappers for simplicity
+
+def conv2d(x, W, b, strides=1):
+
+    # Conv2D wrapper, with bias and relu activation
+
+    x = tf.nn.conv2d(x, W, strides=[1, strides, strides, 1], padding='SAME')
+
+    x = tf.nn.bias_add(x, b)
+
+    return tf.nn.relu(x)
+
+
+
+
+
+def maxpool2d(x, k=2):
+
+    # MaxPool2D wrapper
+
+    return tf.nn.max_pool(x, ksize=[1, k, k, 1], strides=[1, k, k, 1],
+
+                          padding='SAME')
+
+
+
+
+
+# Create model
+
+def conv_net(x, weights, biases):
+
+
+    x = tf.reshape(x, shape=[-1, T_HEIGHT, T_WIDTH, T_channels])
+
+
+
+    # Convolution Layer
+
+    conv1 = conv2d(x, weights['wc1'], biases['bc1'])
+
+    # Max Pooling (down-sampling)
+
+    conv1 = maxpool2d(conv1, k=2)
+
+
+
+    # Convolution Layer
+
+    conv2 = conv2d(conv1, weights['wc2'], biases['bc2'])
+
+    # Max Pooling (down-sampling)
+
+    conv2 = maxpool2d(conv2, k=2)
+
+
+
+
+    # Fully connected layer
+
+    # Reshape conv2 output to fit fully connected layer input
+
+    fc1 = tf.reshape(conv2, [-1, weights['wd1'].get_shape().as_list()[0]])
+
+    fc1 = tf.add(tf.matmul(fc1, weights['wd1']), biases['bd1'])
+
+    fc1 = tf.nn.relu(fc1)
+
+    # Output, class prediction
+
+    out = tf.add(tf.matmul(fc1, weights['out']), biases['out'])
+
+    return out
+
+
+
+# Store layers weight & bias
+
+weights = {
+
+    # 5x5 conv, 1 input, 32 outputs
+
+    'wc1': tf.Variable(tf.random_normal([3, 3, 3, 32])),
+
+    # 5x5 conv, 32 inputs, 64 outputs
+
+    'wc2': tf.Variable(tf.random_normal([3, 3, 32, 64])),
+
+    # fully connected, 32*32*64 inputs, 1024 outputs
+
+    'wd1': tf.Variable(tf.random_normal([8*8*64, 512])),
+
+    # 1024 inputs, 10 outputs (class prediction)
+
+    'out': tf.Variable(tf.random_normal([512, num_classes]))
+
+}
+
+
+
+biases = {
+
+    'bc1': tf.Variable(tf.random_normal([32])),
+
+    'bc2': tf.Variable(tf.random_normal([64])),
+
+    'bd1': tf.Variable(tf.random_normal([512])),
+
+    'out': tf.Variable(tf.random_normal([num_classes]))
+
+}
+
+
+
+# Construct model
+
+logits = conv_net(X, weights, biases)
+
+prediction = tf.nn.softmax(logits)
+
+
+
+# Define loss and optimizer
+
+loss_op = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(
+                         logits=logits, labels=Y))
+
+optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
+
+train_op = optimizer.minimize(loss_op)
+
+
+
+
+
+# Evaluate model
+
+correct_pred = tf.equal(tf.argmax(prediction, 1), tf.argmax(Y, 1))
+
+accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.int64))
+
+
+
+# Initialize the variables (i.e. assign their default value)
+tfrec_dev_input_fn = trec.decode_tfrecord.tfrec_data_input_fn(["catvdog.tfrecords"],batch_size=batch_size)
+features, labels = tfrec_dev_input_fn()
+
+init = tf.global_variables_initializer()
+
+
+
+# Start training
+
+with tf.Session() as sess:
+
+
+
+    # Run the initializer
+
+    sess.run(init)
+
+
+
+    for step in range(1, num_steps+1):
+        img, label = sess.run([features['image'], labels])
+        
+        # Run optimization op (backprop)
+        sess.run(train_op, feed_dict={X: img, Y: label})
+        
+        if step % display_step == 0 or step == 1:
+            # Calculate batch loss and accuracy
+            loss, acc = sess.run([loss_op, accuracy], feed_dict={X: img,
+                                                                 Y: label})
+            print("Step " + str(step) + ", Minibatch Loss= " + \
+                  "{:.4f}".format(loss) + ", Training Accuracy= " + \
+                  "{:.3f}".format(acc))
+
+    print("Optimization Finished!")
+    '''tfrec_dev_input_fn = trec.decode_tfrecord.tfrec_data_input_fn(["catvdogtest.tfrecords"], batch_size=1000)
+    features, labels = tfrec_dev_input_fn()
+    img, label = sess.run([features['image'], labels])
+    print("Testing Accuracy:", \
+            sess.run(accuracy, feed_dict={X: img,
+                                      Y: label}))'''
+    sess.close()
\ No newline at end of file
diff --git a/Work/NN-Xray.py b/Work/NN-Xray.py
new file mode 100644
index 0000000..40788bf
--- /dev/null
+++ b/Work/NN-Xray.py
@@ -0,0 +1,112 @@
+import tensorflow as tf
+import numpy as np
+import tfRecord as trec
+T_HEIGHT = 256
+T_WIDTH = 256
+
+TRAIN_FILE = 'tftrain.tfrecords'
+TEST_FILE = 'tftest.tfrecords'
+
+learning_rate = 0.001
+num_steps = 1500
+batch_size = 16
+display_step = 10
+
+n_hidden_1 = 3 # 1st layer number of neurons
+n_hidden_2 = 3 # 2nd layer number of neurons
+num_input = 3072 #  data input (img shape: 256*256)
+num_classes = 2 # 
+
+# tf Graph input
+X = tf.placeholder("float", [None, num_input])
+Y = tf.placeholder(tf.int64, [None, num_classes])
+
+# Store layers weight & bias
+weights = {
+    'h1': tf.Variable(tf.random_normal([num_input, n_hidden_1])),
+    'h2': tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
+    'out': tf.Variable(tf.random_normal([n_hidden_2, num_classes]))
+}
+biases = {
+    'b1': tf.Variable(tf.random_normal([n_hidden_1])),
+    'b2': tf.Variable(tf.random_normal([n_hidden_2])),
+    'out': tf.Variable(tf.random_normal([num_classes]))
+}
+
+
+# Create model
+def neural_net(x):
+    # Hidden fully connected layer with 256 neurons
+    layer_1 = tf.add(tf.matmul(x, weights['h1']), biases['b1'])
+    z1 = tf.nn.relu(layer_1)
+    # Hidden fully connected layer with 256 neurons
+    layer_2 = tf.add(tf.matmul(z1, weights['h2']), biases['b2'])
+    # Output fully connected layer with a neuron for each class
+    '''z2 = tf.nn.relu(layer_2)'''
+    out_layer = tf.matmul(layer_2, weights['out']) + biases['out']
+    return out_layer
+
+# Construct model
+logits = neural_net(X)
+prediction = tf.nn.softmax(logits)
+
+# Define loss and optimizer
+loss_op = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(
+   labels=Y, logits=logits))
+optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
+train_op = optimizer.minimize(loss_op)
+
+# Evaluate model
+correct_pred = tf.equal(tf.argmax(prediction,1), tf.argmax(Y, 1))
+accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
+####
+
+
+
+
+
+tfrec_dev_input_fn = trec.decode_tfrecord.tfrec_data_input_fn(["catvdog.tfrecords"])
+features, labels = tfrec_dev_input_fn()
+
+# Initialize the variables (i.e. assign their default value)
+init = tf.global_variables_initializer()
+
+# Start training
+with tf.Session() as sess:
+
+    # Run the initializer
+    sess.run(init)
+    
+    for step in range(1, num_steps+1):
+        img, label = sess.run([features['image'], labels])
+        
+        # Run optimization op (backprop)
+        sess.run(train_op, feed_dict={X: img, Y: label})
+        
+        if step % display_step == 0 or step == 1:
+            # Calculate batch loss and accuracy
+            loss, acc = sess.run([loss_op, accuracy], feed_dict={X: img,
+                                                                 Y: label})
+            print("Step " + str(step) + ", Minibatch Loss= " + \
+                  "{:.4f}".format(loss) + ", Training Accuracy= " + \
+                  "{:.3f}".format(acc))
+
+    print("Optimization Finished!")
+    tfrec_dev_input_fn = trec.decode_tfrecord.tfrec_data_input_fn(["catvdogtest.tfrecords"], batch_size=1000)
+    features, labels = tfrec_dev_input_fn()
+    img, label = sess.run([features['image'], labels])
+    print("Testing Accuracy:", \
+            sess.run(accuracy, feed_dict={X: img,
+                                      Y: label}))
+    '''tfrec_dcatvdog= trec.decode_tfrecord.tfrec_data_catvdog(["catvdogtest2.tfrecords"])
+    a = tfrec_dcatvdog()
+    for i in range(750):
+        img = sess.run([a['image']])
+        print("prediction :", \
+                sess.run(prediction, feed_dict={X:img}))
+    print(sess.run([weights['h1'], biases['b1'], weights['h2'], biases['b2']]))
+    print(sess.run([weights['out'],biases['out']]))'''
+    
+
+
+    
\ No newline at end of file
diff --git a/Work/numpyrec/NN.py b/Work/numpyrec/NN.py
new file mode 100644
index 0000000..302af5d
--- /dev/null
+++ b/Work/numpyrec/NN.py
@@ -0,0 +1,100 @@
+import tensorflow as tf
+import numpy as np
+T_HEIGHT = 256
+T_WIDTH = 256
+
+TRAIN_FILE = 'tftrain.tfrecords'
+TEST_FILE = 'tftest.tfrecords'
+
+learning_rate = 0.001
+num_steps = 1500
+batch_size = 16
+display_step = 10
+
+n_hidden_1 = 3 # 1st layer number of neurons
+n_hidden_2 = 3 # 2nd layer number of neurons
+num_input = 49152 #  data input (img shape: 256*256)
+num_classes = 2 # 
+
+# tf Graph input
+X = tf.placeholder("float", [None, num_input])
+Y = tf.placeholder(tf.int64, [None, num_classes])
+
+# Store layers weight & bias
+weights = {
+    'h1': tf.Variable(tf.random_normal([num_input, n_hidden_1])),
+    'h2': tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
+    'out': tf.Variable(tf.random_normal([n_hidden_2, num_classes]))
+}
+biases = {
+    'b1': tf.Variable(tf.random_normal([n_hidden_1])),
+    'b2': tf.Variable(tf.random_normal([n_hidden_2])),
+    'out': tf.Variable(tf.random_normal([num_classes]))
+}
+
+
+# Create model
+def neural_net(x):
+    # Hidden fully connected layer with 256 neurons
+    layer_1 = tf.add(tf.matmul(x, weights['h1']), biases['b1'])
+    z1 = tf.nn.relu(layer_1)
+    # Hidden fully connected layer with 256 neurons
+    layer_2 = tf.add(tf.matmul(z1, weights['h2']), biases['b2'])
+    # Output fully connected layer with a neuron for each class
+    '''z2 = tf.nn.relu(layer_2)'''
+    out_layer = tf.matmul(layer_2, weights['out']) + biases['out']
+    return out_layer
+
+# Construct model
+logits = neural_net(X)
+prediction = tf.nn.softmax(logits)
+
+# Define loss and optimizer
+loss_op = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(
+   labels=Y, logits=logits))
+optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
+train_op = optimizer.minimize(loss_op)
+
+# Evaluate model
+correct_pred = tf.equal(prediction, tf.argmax(Y, 1))
+accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
+####
+
+
+# Initialize the variables (i.e. assign their default value)
+init = tf.global_variables_initializer()
+
+# Start training
+with tf.Session() as sess:
+
+    # Run the initializer
+    sess.run(init)
+    
+    for step in range(1, num_steps+1):
+        img, label = sess.run([features['image'], labels])
+        
+        # Run optimization op (backprop)
+        sess.run(train_op, feed_dict={X: img, Y: label})
+        
+        if step % display_step == 0 or step == 1:
+            # Calculate batch loss and accuracy
+            loss, acc = sess.run([loss_op, accuracy], feed_dict={X: img,
+                                                                 Y: label})
+            print("Step " + str(step) + ", Minibatch Loss= " + \
+                  "{:.4f}".format(loss) + ", Training Accuracy= " + \
+                  "{:.3f}".format(acc))
+
+    print("Optimization Finished!")
+    print("Testing Accuracy:", \
+            sess.run(accuracy, feed_dict={X: img,
+                                      Y: label}))
+    for i in range(750):
+        img = sess.run([a['image']])
+        print("prediction :", \
+                sess.run(prediction, feed_dict={X:img}))
+    print(sess.run([weights['h1'], biases['b1'], weights['h2'], biases['b2']]))
+    print(sess.run([weights['out'],biases['out']]))
+    
+
+
+    
\ No newline at end of file
diff --git a/Work/numpyrec/nprecord.py b/Work/numpyrec/nprecord.py
new file mode 100644
index 0000000..fe99af1
--- /dev/null
+++ b/Work/numpyrec/nprecord.py
@@ -0,0 +1,64 @@
+import os 
+import numpy as np
+from glob import glob
+import cv2 as cv
+###############
+# Cat vs Dog #
+###############
+
+IMAGE_CHANNELS = 3
+IMAGE_HEIGHT = 128
+IMAGE_WIDTH = 128
+NUM_CLASS = 2
+gen_file_path = "train/train/"
+
+def get_paths(ur_path):
+    file_class1 = os.path.join(gen_file_path,'cat*.jpg' )
+    file_class2 = os.path.join(gen_file_path, 'dog*.jpg')
+    cat_files = glob(file_class1)
+    dog_file = glob(file_class2)
+    size = len(cat_files) + len(dog_file)
+    return cat_files, dog_file, size
+
+def get_numpy(class1_files, class2_files, size):
+    dataX = np.zeros((size,IMAGE_HEIGHT, IMAGE_WIDTH,IMAGE_CHANNELS), dtype=np.float64)
+    dataY = np.zeros((size))
+    i = 0
+    for file in class1_files:
+            img = cv.imread(file)
+            img = cv.resize(img,(IMAGE_HEIGHT, IMAGE_WIDTH))
+            img_np = np.array(img)
+            dataX[i] = img_np
+            dataY[i]= 0
+            i += 1
+    for file in class2_files:
+        try:
+            img = cv.imread(file)
+            img = cv.resize(img, (IMAGE_HEIGHT, IMAGE_WIDTH))
+            img_np = np.array(img)
+            dataX[i] = img_np
+            dataY[i] = 1
+            i +=1
+        except:
+            continue
+    return dataX, dataY
+
+def make_npfile(dir, inputs, labels):
+    ######
+    #Space dedicated for pre-processing methods
+    ######
+    if  (dir == ""):
+        dir = os.getcwd()
+    np.save(dir, [inputs, labels], ["inputs, labels"])
+    return 0
+
+def load_npfile(filename):
+    f = os.path.join(os.path.dirname(filename), filename)
+    data = np.load(f)
+    features = data[0]
+    labels = data[1]
+    return features, labels
+alpha, beta, gamma = get_paths(gen_file_path)
+a , b = get_numpy(alpha, beta, gamma)
+make_npfile("", a, b)
+'''a, b= load_npfile("train.npy")'''
diff --git a/Work/tfRecord/Tfrecord.py b/Work/tfRecord/Tfrecord.py
new file mode 100644
index 0000000..3c8fd3d
--- /dev/null
+++ b/Work/tfRecord/Tfrecord.py
@@ -0,0 +1,199 @@
+from random import shuffle, seed
+
+import glob
+
+import numpy as np
+
+import cv2
+
+import tensorflow as tf
+
+import sys
+
+import matplotlib.pyplot as plt
+
+
+
+UR_DATA_PATH = ""
+
+shuffle_data = True
+
+
+
+# This function returns the different paths and corresponding labels
+
+def read_paths(ur_path):
+
+    #Reading the path of each data(ex: image) and then extracting the labels from path
+
+    data_path = glob.glob(ur_path)
+
+    labels = [int(i.split('.')[0][-1]) for i in data_path]
+
+    return data_path, labels
+
+
+def shuffling_data(data_path, labels):
+
+    c = list(zip(data_path, labels))
+
+    shuffle(c)
+
+    #NOTE: data and labeld are tuples. Care about immutability 
+
+    data, labeled = zip(*c) #NOTE:ith element in data correpand to ith label elmt
+
+    return data, labeled
+
+
+def split_data(training_prop,test_prop, data, labels, dev_prop=0):
+
+    ''' 
+
+    - training_prop is required .numerical float < 1
+
+    - dev_prop is optional. Numerical float < 0.3
+
+    - test_prop is required. Numerical float < 0.3
+
+    - data and labels are lists 
+
+    '''
+
+    #NOTE: data is a list of paths of each image
+
+    training_data = data[0:int(training_prop*len(data))]
+
+    training_labels = labels[0:int(training_prop*len(labels))]
+
+
+
+    if dev_prop != 0:
+
+        dev_data = data[int(training_prop*len(data)):int((1-dev_prop)*len(data))]
+
+        dev_labels = labels[int(training_prop*len(labels)):int((1-dev_prop)*len(labels))]
+
+        test_data = data[int((1-dev_prop)*len(data)):]
+
+        test_labels = labels[int((1-dev_prop)*len(labels)):]
+
+        return training_data,training_labels, dev_data, dev_labels, test_data, test_labels
+
+    else:
+
+        test_data = data[int((1-test_prop)*len(data)):]
+
+        test_labels = labels[int((1-test_prop)*len(labels)):]
+
+        return training_data, training_labels, test_data, test_labels
+
+
+
+# Load image 
+
+def load_image(data, s_width, s_height):
+
+    #cv2 load data from data (value of one path) 
+    img_load = cv2.imread(data)
+    img_after = cv2.resize(img_load, dsize=(s_width, s_height))
+
+    img_after = cv2.cvtColor(img_after, cv2.COLOR_BGR2GRAY)
+    img_after = img_after.astype(np.float32)
+    img_after = img_after.tostring()
+    return img_after
+
+a = load_image("pulmonary_chest_xray/ChinaSet_AllFiles/CXR_png/CHNCXR_0001_0.png", 1000, 1000)
+
+
+def _int64_feature(value):
+
+    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))
+
+
+
+def _bytes_feature(value):
+
+    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))
+
+
+
+def tfrecord_file(data, labeled, x_filename, s_width, s_height):
+
+    #NOTE:data is a tuple (list) of paths of each image
+
+    #NOTE: data can come from the dev-set or test-set
+
+    #NOTE: Same for labeled
+
+
+
+    file_path = x_filename
+
+    #Open a TFRecordWriter
+
+    writer = tf.python_io.TFRecordWriter(file_path)
+
+    for i in range(1):
+
+        new_img = load_image(data[i], s_width, s_height)
+
+        new_label = labeled[i]
+
+        new_feature = {'train/label': _int64_feature(new_label),
+
+                        'train/image': _bytes_feature(tf.compat.as_bytes(new_img))
+        }
+
+
+        #NOTE: PLease refer to this url for defenirtion: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
+
+        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
+
+        writer.write(example.SerializeToString())
+
+    writer.close()
+
+    sys.stdout.flush()
+
+    return 0
+
+
+data_path, labels = read_paths("pulmonary_chest_xray/ChinaSet_AllFiles/CXR_png/*.png")
+
+tfrecord_file(data_path, labels, "Tfexamp.tfrecords", 1000, 1000)
+
+'''data_test_path, test_labels = read_paths("pulmonary_chest_xray/MontgomerySet/CXR_png/*.png")
+
+tfrecord_file(data_test_path, test_labels, "TfTestSet.tfrecords",1000,1000)'''
+
+
+
+def read_file(x_filename, x_capacity=662):
+
+    feature = {'train/image': tf.FixedLenFeature([], tf.string),
+
+                'train/label': tf.FixedLenFeature([], tf.int64)
+
+    }
+
+    queue = tf.train.string_input_producer([x_filename], num_epochs=1)
+
+    reader = tf.TFRecordReader()
+
+    _, serialized_example = reader.read(queue)
+
+    features = tf.parse_single_example(serialized_example, feature=feature)
+
+    
+
+    #x means training, dev or test
+
+    x_image = tf.decode_raw(features['train/image'], tf.float32)
+
+    x_label = tf.cast(features['train/label'], tf.int64)
+
+
+    #NOTE: In this line, we can add a instruction to preprocess the images to their original shape
+
+    return x_image, x_label
\ No newline at end of file
diff --git a/Work/tfRecord/__init__.py b/Work/tfRecord/__init__.py
new file mode 100644
index 0000000..dab8f1b
--- /dev/null
+++ b/Work/tfRecord/__init__.py
@@ -0,0 +1,2 @@
+from . import Tfrecord
+from . import decode_tfrecord
\ No newline at end of file
diff --git a/Work/tfRecord/__pycache__/Tfrecord.cpython-36.pyc b/Work/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..b256d391818616bea6ef6a8b64a54b8e549d78a9
GIT binary patch
literal 3599
zcmbVPOOM;g5hj}tQItkAqxECPyGVr1Yqde-vAxJ4F^sjx@ooU`Y%nv6O^8_#w7Mlx
zONn%ovS$Vq5J1k+L2}6{e;`19L9PLE&bhBSB>oF~Nxo`I@_05#kWgTEb+N0uy85fC
z)}4(F|IfzXgHPIq@i*hb<AVMWZ~iAV+~8(nOw7QXuz;B~W{DM8y0-&I_io_oz7}}8
zuLpJA`{PE?<W{i3?V!aS?!GXBO<v<3`YnEg*SY_~47PcLH^FI(Hs9ba-db3}TYQsm
zfp>*(^EUdc{4IV3{WY=4uP)3J7F-W@#5HkU?C@(^#(`~b^Bukmjk}y38QphY!G5F5
zR4pIP27^Q>I~Rf<DC4+q(kDC?UWRucZ{9&u7-OR}7UtaKCTA~<(tJoGR+=RnkXGwp
zElh$A-XpyEyJ&_G{o=;lEFd;EIRv@6YjA59)Nk2S8{)Pydy-avrJH}^frVB)EtHdI
zNnBvWdV9)@|6s@_=6!Xi|D%)7av^iw-nxDAP~=}0X?Aif^5T=nC&l1U^i#=Cjt5c`
zz3l8O|85fJML(Tn_r6Kik*Th6{tYQ29%fN7%6oT{v>zq;JrJv@T~pb^Bz@Agg4#^#
zd8*E%BC1F?iJpig*Jf4C2GlD5wbET$i|}LEK5H=vXc>K4n;vqNjaU2neGr9Fm?d1b
zVC3O~mFP|4v0ob}Y;Mi%(kgB3b@QNWDVq#c)_I&s(%E(77CO0&MwxwOlclPrjTHQ{
zm8}W{@pKr{yl+D&_s|%y)3cjQ?tt!-jr26|(mu4UJ2<M6aXMCQ-NL+r1N{}-HReuf
zt@+jkzFXSd{L+$tE!mq=1>&c)cP{IxmDY0JQp<&T7xj9i@3IDNX<O=~WBqL@wfyFn
zPU)7l(kttwKX8w*A3fK+Xvrl#)wh4AhjakU+#*1q|E1&m9Xvni6f%mZ<l#)FStrgr
zQaqi-Qt(c1I-3X?18|)|l14@6Zs*oYii>A!fix@PbQ&eS&cW9N_Ih_#A_btomRT+G
zUu5XiJJFPPDin7jDLMqGd}+RpN=b))&}EXKp{ztqyNzJujeQ06_3H8(W@-E(@hx|8
zVK%K8Y_5f?b*q{BlTR{BZlzSS8(M$Gzl>sE=Ln79TUa|M1B|U5wqtwjJ^bFmZ_B1u
z-oTeW<u`iXh`k43AkQK85O=`Lm?6$g%5U6!$>ugVh&#0Q0;AF%JB2&0VJ^|$!k``}
z&knjK3P2v8i*8+6{b#q88&4=1C|3$%2VVbKv7aWX3|jl2KK}GDeDKlX?T-%c|3bM@
zUYun@v1z7i8WMMIE3ZiNLIOYtZQ2<!clbQ!#YlO1I1=%2R0Li;8HRum59&lkgF?u!
zd5OM<@gy3G5NdMDqlV?1ZPqfIre{iu{__vBSuz3G<XPAs5ui?Gbar!p6i=fgQH1xC
z<WZc6{O11855sIayt)7J0g-!qd$+>99yswo;NEpLpnSegZ{&P(@fcx-)+&u5YQamC
zl$WM1B1pB)vnZJf5I(a14?hSAjHnN+7$`@BM%5mPsF+Ei>OZ9!k7<G_zXRQuciajq
z9fL|>f+;%Z8%%y5^#5(7UW`n9at2f`O_bjw3*Muqi$>KS(8vF1T_wJ>E{D*{x*q6t
z)}?bIr0T5u8xSR%JCz$t2l#VOM+Nfj0yX)IlX<PMQ9WwA233iLHTQsFjOO*iDeDWf
za3NV6d!@I41ZTfzb03`gxM~}v&n+B?{nDH_N_X5W8{>@yBEmGD-uT=Ar?oIpjW)+y
z<87k2gM)JqbcX1%^M*d6n>tKa1C2O%tG3)INN=@gonE?a?K@S^&Wcew4dYa`jvrO^
z`*Rrs%F2<^bSN}J_0dW4kaDS%M}_jDJk-9_?pL*+oJNxjHyiI!b?~`zo|7K6eS~xy
zB?!Xf^hn2)a{5V{3uWbJxpMP@r!(9+2FWZRb(>WtyG~2_r{Ms&pGHX7?}8z}Mcg{^
zme&Bif|OJvgW6O)N2Co^oe1rXph4tv->N1m6E7onqgY;7^lE3&l~V;ad>h5M#kQEo
zTz19XVQo-8lZ1)$5B|rHeV@YgeEWEyjp(hG%wPQu7l`%6#E0s2<o*%^g4~c=w$Smb
zX={P>2oMnmDN#o>1B7r<0mrS<;uwXTuBUTnPVn<kbQvJV09HV+tZ`?T^!iAw9{L72
zuI51dyiVm=Uw3LKL2rO+HP8;OAc&wuQ4K|w_D4C=!cB6}5|tW=tAUom5mVHv%}4R6
z;E%<02^c~GWnAUx1X983YQ;lN$!5CF@>w6yTtoE`IUZwM=lUyIBy$nwPy&Q3Pg|k3
z^;0fF89kRBTG;L*<9B^|4}$<0>qlACkBc*Hw{o7&#7t1#(3;IKXyeLT0hq6p$~}^;
zomQEhh{_$5HY#}O`)Hk%XjJ_XA9A{b8m!6MHX*6ST5g-Q9eD^Yyhd3Y-9zebV0~N=
z(uuxaN+fqfZlX~R9G#s>x=G8oiQyMwnx_(V>v@$7e@xu%NmOJ>S|srk+`)-b?y|SC
zqzIay9fl9@AKwpucK`U}pt<~4qKmn{DeEh!uG>2Nl^MOx02=?V@4MRk3n8Ulpm*=x
nouqu0h<miN{JUr<lvY1P7qeZj;cfUmuhFzT7r$<kZF&C!%5slT

literal 0
HcmV?d00001

diff --git a/Work/tfRecord/__pycache__/__init__.cpython-36.pyc b/Work/tfRecord/__pycache__/__init__.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..14365e91ca043da0883b22aa9a2927f61aeaa8eb
GIT binary patch
literal 205
zcmXr!<>flW_&my#fq~&M5W@izkmUfx#WFx5g&~D8has0Sijfh-W&*OAqL_ee=3oX*
zmY0k`C7O)4I6~5jQj_zGQh;p!6d)@lHNFHY<fqAei=hZ)Xc04zSjkYt3SxtZU&hH+
zF`>n&Ma5v!&?v?wwK%&ZzaS<gwYbFBC#EDV2yARje0*kJW=VX!UP0w84x8Nkl+v73
MJCNnYAcJ`r0bnUIT>t<8

literal 0
HcmV?d00001

diff --git a/Work/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc b/Work/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..34952f566042af9c4999d1a4f01997fd772b8c41
GIT binary patch
literal 3051
zcmbtWTW{1x6!wg->ufGSproaAd!emcSZIK%LJ+D#^Uy*>3l-85WSpJZT`#+~JD!9j
z+I<RAcxe7X|3LrJJXPvLmA5|kr9Efty-=kB30re!d^|pL=FB<Y=gYIRe(jq-8^(fR
z{9#O8CXBy^mOlW&3`TlJgXo?%Ox;@zOZRryX}EgiH7cyiEM`9^4WG?2mw7OnVRNj)
zd>B>v44c^_1gzaRg84(RY^)s;kcY6m(Zn_JvY_D@TfPWl-`FSnv}H2F=yPvq4ovvo
zTx-HBec2!>v=h+sH4u8DF*KynC0+W0?2#cEkgho}%8~VgmLvPcNaySsLpq>cch6X-
zLvv^i%z@RS#Zu@cUJf)|?mo198w4)VMoAetpyMa-^?)+^D;*NhSp#E0_UI7b+XKqX
zCH%qxjS0OA%>%8awUs@}Y~~!$p`{~e4DEqEFu`4u!Omk+I#pQ<h_ad7$#_7O9rvR)
zS9UMj;5~sJ3Uu`l!=mT$Om>>1U7!}Dt9S7>48Q|2(7S-utwn>Grwt4lY~J3zaw|=?
zxRA9cdXXo!mPq@xN0-jmmakrYT)VJRdvxJa{p$G|h{v^7FO6hgZ-Z61{<?Yi4|L_z
z?>E~E`pRwwrlPW?tY(x;<?4?vF9)>Y>NUb`y@%fnF$-V+eXV(I{imFZT+1((*KhOu
zsZ6u=2RxT|@2tz#kGz=*w$7kqJd`b=pX=F<^1~z*{iql3a@=W-*=5jEWcLR>D`b(3
z!Z^t`W!OsU*L!I*>g6l-qf}WWavqkZCo{+P17G-HT+Oe>+nn9u$tsWJrr^rSrHGR@
zoKhmcyrLEhLpi8=JmlL^KkISjm(GjmnX)%w=2yy16CQR_DQ3_teho2lR7JK*FO=6R
zmdKS;_@ydE0KtmR05K1r>L?as4lT`w(_6vf%TX9~jx5qDJx<P0&-Uo<jW-HW;1%Jp
zFvM#9!D?|xx8W)0QknfIL$5Q~fwDIu+3cuO{pcwVao0I)JB%e4kxYg1+FXVS-<E+b
z&fpGQw24-{(HyaadPf&P4%y{vi(wAut}3m#$CIefbLA(S{g7vAvy)E+q%uBk3|A2h
z;w@H*Pkm?}snXq#UoRNiYG8>^!H@`0e1hUL6qw9`Stus&^^u%7!a7bb^L3#nK8L5n
z8kB7l%NMT-G)(7P(DGkEU>SGm0PydCXs|K67NDBBoeb!0<8ccy4FJeO5S9*%-LA*z
zlF_YneFP-{8GEb;IGWOrX4{~hDP|Ad(t<OYIrM%$`w-v*fWt+Ava(2al+)pHyCVhS
zkFuY|Om-$16;z8fsbo8{lO|!Ds^jAXpQL~-MHDB>{kANZ`FmK6s{lp!i{dh&Jc=Z6
z%c4-euD7rk1IE}}QEwBfY-30AT&seL^0&iLYT&pTWl=MhJB|6Gm?o>cWtBEuC}k`@
zQHQ;Viy#_)+;4|Pg%>4<=Qf>Of=5iRvkWit(;y5ohlvIE0`({~bB;Rnv?*@DyQ7H^
z3v}OTk^dt>OoTBB;zIu+!4b_(kzh%9Hm1mkQ%4l>648C9&0%(xR7VeJ44euYFgK<)
z3rLJ;60?!on8~c?2IL%M9cC6I13T7~vZhZzAOf8pvl1@-YF07^KwEFjN&;)=C|1(B
zsPR@*VG8)D$@JwT)u8$iUQgKM-c<cIBe})W_8XFkQ@DcXE{>tV-y4DdL*gWez!8{9
zI^zX0n|KEW;@5~olOhSV?+7v)I@e&3<JJtSvEBrvw|1|;iTmIR@}77X#aR^Zq4)sB
zhadnA5wyieD9)iktcK8P44k-!7<m_3;U4i$0{+?_0U|p0TJF6Pddn3x-~z@#85i_l
z9@zJ|Hm0OOtM7{y=y%_$Uo2k&YRJ0&27Wa1fC?-&f%ys>Igj16YF7D@CwVI1&*oWy
zZ#FOw{EyHiQ#8%hv3?}8UMhR>2K=VrNI7M1W<8+C=GL;J%`J`l8W*%jQ%5=uobs;s
bQ?}XTE9g^x90WuSZl8Frf6ia>18?SETyx}{

literal 0
HcmV?d00001

diff --git a/Work/tfRecord/decode_tfrecord.py b/Work/tfRecord/decode_tfrecord.py
new file mode 100644
index 0000000..9dd3828
--- /dev/null
+++ b/Work/tfRecord/decode_tfrecord.py
@@ -0,0 +1,128 @@
+#NOTE: refer to this url: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/how_tos/reading_data/fully_connected_reader.py
+import tensorflow as tf
+import numpy as np
+import os
+import matplotlib.pyplot as plt
+import cv2 as cv
+
+
+
+
+
+
+def tfrec_data_input_fn(filenames, num_epochs=1, batch_size=16):
+    
+    def _input_fn():
+        def _parse_record(tf_record):
+            features = {
+                'image': tf.FixedLenFeature([], dtype=tf.string),
+                'label': tf.FixedLenFeature([], dtype=tf.int64)
+            }
+            record = tf.parse_single_example(tf_record, features)
+
+            image_raw = tf.decode_raw(record['image'], tf.uint8)
+            label = tf.one_hot(tf.cast(record['label'], tf.int32), depth=2)
+            
+            def _normalize(image_x):
+                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+                image = tf.cast(image_x, tf.float32) * (1. / 255) - 0.5
+                return image
+            
+            image_raw = _normalize(image_raw)
+            '''image_raw = tf.reshape(image_raw, (32,32,3))'''#For vizualizing 
+            return { 'image': image_raw }, label
+        
+        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
+        dataset = tf.data.TFRecordDataset(filenames)
+        dataset = dataset.map(_parse_record)
+
+        dataset = dataset.repeat(num_epochs)
+        dataset = dataset.batch(batch_size)
+
+        iterator = dataset.make_one_shot_iterator()
+        features, labels = iterator.get_next()
+
+        return features, labels
+    
+    return _input_fn
+
+
+def read_file(x_filename, x_capacity=800):
+    record_iterator = tf.python_io.tf_record_iterator(path=x_filename)
+
+    example = tf.train.Example()
+    for str_rec in record_iterator:
+        example.ParseFromString(str_rec)
+        height = int(example.features.feature['height']
+                                 .int64_list
+                                 .value[0])
+
+        width = int(example.features.feature['width']
+                                .int64_list
+                                .value[0])
+    
+
+        img_string = (example.features.feature['image']
+                                  .bytes_list
+                                  .value[0])
+
+        label = (example.features.feature['label'].int64_list.value[0])
+
+
+    return 0
+
+
+def tfrec_data_catvdog(filenames, num_epochs=1, batch_size=16):
+    
+    def _input_fn():
+        def _parse_record(tf_record):
+            features = {
+                'image': tf.FixedLenFeature([], dtype=tf.string),
+            }
+            record = tf.parse_single_example(tf_record, features)
+
+            image_raw = tf.decode_raw(record['image'], tf.uint8)
+            
+            
+            def _normalize(image):
+                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+                image = tf.cast(image, tf.float32) * (1. / 255) - 0.5
+                return image
+
+            image_n = _normalize(image_raw)
+            
+            
+
+            return { 'image': image_n }
+        
+        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
+        dataset = tf.data.TFRecordDataset(filenames)
+        dataset = dataset.map(_parse_record)
+
+        '''dataset = dataset.repeat(num_epochs)'''
+        '''dataset = dataset.batch(batch_size)'''
+
+        iterator = dataset.make_one_shot_iterator()
+        features= iterator.get_next()
+
+        return features
+    
+    return _input_fn
+
+
+
+
+
+read_file("./catvdogtest.tfrecords")
+
+'''tfrec_dev_input_fn = tfrec_data_input_fn(["catvdog.tfrecords"])
+features, labels = tfrec_dev_input_fn()
+
+with tf.Session() as sess:
+    img, label = sess.run([features['image'], labels])
+    img = np.reshape(img, (img.shape[0], -1))
+    print(img.shape)
+    for i in range(16):
+        plt.imshow(img[i])
+        plt.show()'''
+        
\ No newline at end of file
diff --git a/Work/utils_data/norma/norma.py b/Work/utils_data/norma/norma.py
new file mode 100644
index 0000000..a0c0fb6
--- /dev/null
+++ b/Work/utils_data/norma/norma.py
@@ -0,0 +1,7 @@
+import tensorflow as tf
+
+
+def normalize(image, label):
+  """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+  image = tf.cast(image, tf.float32) * (1. / 255) - 0.5
+  return image, label
\ No newline at end of file

From 88438908b5bfe76a334b15da8cce171cece3c639 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:00:04 +0200
Subject: [PATCH 02/28] Delete Tfrecord.py

---
 Work/tfRecord/Tfrecord.py | 199 --------------------------------------
 1 file changed, 199 deletions(-)
 delete mode 100644 Work/tfRecord/Tfrecord.py

diff --git a/Work/tfRecord/Tfrecord.py b/Work/tfRecord/Tfrecord.py
deleted file mode 100644
index 3c8fd3d..0000000
--- a/Work/tfRecord/Tfrecord.py
+++ /dev/null
@@ -1,199 +0,0 @@
-from random import shuffle, seed
-
-import glob
-
-import numpy as np
-
-import cv2
-
-import tensorflow as tf
-
-import sys
-
-import matplotlib.pyplot as plt
-
-
-
-UR_DATA_PATH = ""
-
-shuffle_data = True
-
-
-
-# This function returns the different paths and corresponding labels
-
-def read_paths(ur_path):
-
-    #Reading the path of each data(ex: image) and then extracting the labels from path
-
-    data_path = glob.glob(ur_path)
-
-    labels = [int(i.split('.')[0][-1]) for i in data_path]
-
-    return data_path, labels
-
-
-def shuffling_data(data_path, labels):
-
-    c = list(zip(data_path, labels))
-
-    shuffle(c)
-
-    #NOTE: data and labeld are tuples. Care about immutability 
-
-    data, labeled = zip(*c) #NOTE:ith element in data correpand to ith label elmt
-
-    return data, labeled
-
-
-def split_data(training_prop,test_prop, data, labels, dev_prop=0):
-
-    ''' 
-
-    - training_prop is required .numerical float < 1
-
-    - dev_prop is optional. Numerical float < 0.3
-
-    - test_prop is required. Numerical float < 0.3
-
-    - data and labels are lists 
-
-    '''
-
-    #NOTE: data is a list of paths of each image
-
-    training_data = data[0:int(training_prop*len(data))]
-
-    training_labels = labels[0:int(training_prop*len(labels))]
-
-
-
-    if dev_prop != 0:
-
-        dev_data = data[int(training_prop*len(data)):int((1-dev_prop)*len(data))]
-
-        dev_labels = labels[int(training_prop*len(labels)):int((1-dev_prop)*len(labels))]
-
-        test_data = data[int((1-dev_prop)*len(data)):]
-
-        test_labels = labels[int((1-dev_prop)*len(labels)):]
-
-        return training_data,training_labels, dev_data, dev_labels, test_data, test_labels
-
-    else:
-
-        test_data = data[int((1-test_prop)*len(data)):]
-
-        test_labels = labels[int((1-test_prop)*len(labels)):]
-
-        return training_data, training_labels, test_data, test_labels
-
-
-
-# Load image 
-
-def load_image(data, s_width, s_height):
-
-    #cv2 load data from data (value of one path) 
-    img_load = cv2.imread(data)
-    img_after = cv2.resize(img_load, dsize=(s_width, s_height))
-
-    img_after = cv2.cvtColor(img_after, cv2.COLOR_BGR2GRAY)
-    img_after = img_after.astype(np.float32)
-    img_after = img_after.tostring()
-    return img_after
-
-a = load_image("pulmonary_chest_xray/ChinaSet_AllFiles/CXR_png/CHNCXR_0001_0.png", 1000, 1000)
-
-
-def _int64_feature(value):
-
-    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))
-
-
-
-def _bytes_feature(value):
-
-    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))
-
-
-
-def tfrecord_file(data, labeled, x_filename, s_width, s_height):
-
-    #NOTE:data is a tuple (list) of paths of each image
-
-    #NOTE: data can come from the dev-set or test-set
-
-    #NOTE: Same for labeled
-
-
-
-    file_path = x_filename
-
-    #Open a TFRecordWriter
-
-    writer = tf.python_io.TFRecordWriter(file_path)
-
-    for i in range(1):
-
-        new_img = load_image(data[i], s_width, s_height)
-
-        new_label = labeled[i]
-
-        new_feature = {'train/label': _int64_feature(new_label),
-
-                        'train/image': _bytes_feature(tf.compat.as_bytes(new_img))
-        }
-
-
-        #NOTE: PLease refer to this url for defenirtion: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
-
-        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
-
-        writer.write(example.SerializeToString())
-
-    writer.close()
-
-    sys.stdout.flush()
-
-    return 0
-
-
-data_path, labels = read_paths("pulmonary_chest_xray/ChinaSet_AllFiles/CXR_png/*.png")
-
-tfrecord_file(data_path, labels, "Tfexamp.tfrecords", 1000, 1000)
-
-'''data_test_path, test_labels = read_paths("pulmonary_chest_xray/MontgomerySet/CXR_png/*.png")
-
-tfrecord_file(data_test_path, test_labels, "TfTestSet.tfrecords",1000,1000)'''
-
-
-
-def read_file(x_filename, x_capacity=662):
-
-    feature = {'train/image': tf.FixedLenFeature([], tf.string),
-
-                'train/label': tf.FixedLenFeature([], tf.int64)
-
-    }
-
-    queue = tf.train.string_input_producer([x_filename], num_epochs=1)
-
-    reader = tf.TFRecordReader()
-
-    _, serialized_example = reader.read(queue)
-
-    features = tf.parse_single_example(serialized_example, feature=feature)
-
-    
-
-    #x means training, dev or test
-
-    x_image = tf.decode_raw(features['train/image'], tf.float32)
-
-    x_label = tf.cast(features['train/label'], tf.int64)
-
-
-    #NOTE: In this line, we can add a instruction to preprocess the images to their original shape
-
-    return x_image, x_label
\ No newline at end of file

From 24f5747bf59c3b860b06ebd635aa5ef8626db9e5 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:00:13 +0200
Subject: [PATCH 03/28] Delete __init__.py

---
 Work/tfRecord/__init__.py | 2 --
 1 file changed, 2 deletions(-)
 delete mode 100644 Work/tfRecord/__init__.py

diff --git a/Work/tfRecord/__init__.py b/Work/tfRecord/__init__.py
deleted file mode 100644
index dab8f1b..0000000
--- a/Work/tfRecord/__init__.py
+++ /dev/null
@@ -1,2 +0,0 @@
-from . import Tfrecord
-from . import decode_tfrecord
\ No newline at end of file

From 025bd34296176dc5dccb9046734b3cce10b13bf7 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:00:21 +0200
Subject: [PATCH 04/28] Delete decode_tfrecord.py

---
 Work/tfRecord/decode_tfrecord.py | 128 -------------------------------
 1 file changed, 128 deletions(-)
 delete mode 100644 Work/tfRecord/decode_tfrecord.py

diff --git a/Work/tfRecord/decode_tfrecord.py b/Work/tfRecord/decode_tfrecord.py
deleted file mode 100644
index 9dd3828..0000000
--- a/Work/tfRecord/decode_tfrecord.py
+++ /dev/null
@@ -1,128 +0,0 @@
-#NOTE: refer to this url: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/how_tos/reading_data/fully_connected_reader.py
-import tensorflow as tf
-import numpy as np
-import os
-import matplotlib.pyplot as plt
-import cv2 as cv
-
-
-
-
-
-
-def tfrec_data_input_fn(filenames, num_epochs=1, batch_size=16):
-    
-    def _input_fn():
-        def _parse_record(tf_record):
-            features = {
-                'image': tf.FixedLenFeature([], dtype=tf.string),
-                'label': tf.FixedLenFeature([], dtype=tf.int64)
-            }
-            record = tf.parse_single_example(tf_record, features)
-
-            image_raw = tf.decode_raw(record['image'], tf.uint8)
-            label = tf.one_hot(tf.cast(record['label'], tf.int32), depth=2)
-            
-            def _normalize(image_x):
-                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
-                image = tf.cast(image_x, tf.float32) * (1. / 255) - 0.5
-                return image
-            
-            image_raw = _normalize(image_raw)
-            '''image_raw = tf.reshape(image_raw, (32,32,3))'''#For vizualizing 
-            return { 'image': image_raw }, label
-        
-        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
-        dataset = tf.data.TFRecordDataset(filenames)
-        dataset = dataset.map(_parse_record)
-
-        dataset = dataset.repeat(num_epochs)
-        dataset = dataset.batch(batch_size)
-
-        iterator = dataset.make_one_shot_iterator()
-        features, labels = iterator.get_next()
-
-        return features, labels
-    
-    return _input_fn
-
-
-def read_file(x_filename, x_capacity=800):
-    record_iterator = tf.python_io.tf_record_iterator(path=x_filename)
-
-    example = tf.train.Example()
-    for str_rec in record_iterator:
-        example.ParseFromString(str_rec)
-        height = int(example.features.feature['height']
-                                 .int64_list
-                                 .value[0])
-
-        width = int(example.features.feature['width']
-                                .int64_list
-                                .value[0])
-    
-
-        img_string = (example.features.feature['image']
-                                  .bytes_list
-                                  .value[0])
-
-        label = (example.features.feature['label'].int64_list.value[0])
-
-
-    return 0
-
-
-def tfrec_data_catvdog(filenames, num_epochs=1, batch_size=16):
-    
-    def _input_fn():
-        def _parse_record(tf_record):
-            features = {
-                'image': tf.FixedLenFeature([], dtype=tf.string),
-            }
-            record = tf.parse_single_example(tf_record, features)
-
-            image_raw = tf.decode_raw(record['image'], tf.uint8)
-            
-            
-            def _normalize(image):
-                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
-                image = tf.cast(image, tf.float32) * (1. / 255) - 0.5
-                return image
-
-            image_n = _normalize(image_raw)
-            
-            
-
-            return { 'image': image_n }
-        
-        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
-        dataset = tf.data.TFRecordDataset(filenames)
-        dataset = dataset.map(_parse_record)
-
-        '''dataset = dataset.repeat(num_epochs)'''
-        '''dataset = dataset.batch(batch_size)'''
-
-        iterator = dataset.make_one_shot_iterator()
-        features= iterator.get_next()
-
-        return features
-    
-    return _input_fn
-
-
-
-
-
-read_file("./catvdogtest.tfrecords")
-
-'''tfrec_dev_input_fn = tfrec_data_input_fn(["catvdog.tfrecords"])
-features, labels = tfrec_dev_input_fn()
-
-with tf.Session() as sess:
-    img, label = sess.run([features['image'], labels])
-    img = np.reshape(img, (img.shape[0], -1))
-    print(img.shape)
-    for i in range(16):
-        plt.imshow(img[i])
-        plt.show()'''
-        
\ No newline at end of file

From 1d3eb9c538513f1c69f3c7cd19aab1ace4044a8e Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:00:31 +0200
Subject: [PATCH 05/28] Delete Tfrecord.cpython-36.pyc

---
 .../tfRecord/__pycache__/Tfrecord.cpython-36.pyc | Bin 3599 -> 0 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)
 delete mode 100644 Work/tfRecord/__pycache__/Tfrecord.cpython-36.pyc

diff --git a/Work/tfRecord/__pycache__/Tfrecord.cpython-36.pyc b/Work/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
deleted file mode 100644
index b256d391818616bea6ef6a8b64a54b8e549d78a9..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 3599
zcmbVPOOM;g5hj}tQItkAqxECPyGVr1Yqde-vAxJ4F^sjx@ooU`Y%nv6O^8_#w7Mlx
zONn%ovS$Vq5J1k+L2}6{e;`19L9PLE&bhBSB>oF~Nxo`I@_05#kWgTEb+N0uy85fC
z)}4(F|IfzXgHPIq@i*hb<AVMWZ~iAV+~8(nOw7QXuz;B~W{DM8y0-&I_io_oz7}}8
zuLpJA`{PE?<W{i3?V!aS?!GXBO<v<3`YnEg*SY_~47PcLH^FI(Hs9ba-db3}TYQsm
zfp>*(^EUdc{4IV3{WY=4uP)3J7F-W@#5HkU?C@(^#(`~b^Bukmjk}y38QphY!G5F5
zR4pIP27^Q>I~Rf<DC4+q(kDC?UWRucZ{9&u7-OR}7UtaKCTA~<(tJoGR+=RnkXGwp
zElh$A-XpyEyJ&_G{o=;lEFd;EIRv@6YjA59)Nk2S8{)Pydy-avrJH}^frVB)EtHdI
zNnBvWdV9)@|6s@_=6!Xi|D%)7av^iw-nxDAP~=}0X?Aif^5T=nC&l1U^i#=Cjt5c`
zz3l8O|85fJML(Tn_r6Kik*Th6{tYQ29%fN7%6oT{v>zq;JrJv@T~pb^Bz@Agg4#^#
zd8*E%BC1F?iJpig*Jf4C2GlD5wbET$i|}LEK5H=vXc>K4n;vqNjaU2neGr9Fm?d1b
zVC3O~mFP|4v0ob}Y;Mi%(kgB3b@QNWDVq#c)_I&s(%E(77CO0&MwxwOlclPrjTHQ{
zm8}W{@pKr{yl+D&_s|%y)3cjQ?tt!-jr26|(mu4UJ2<M6aXMCQ-NL+r1N{}-HReuf
zt@+jkzFXSd{L+$tE!mq=1>&c)cP{IxmDY0JQp<&T7xj9i@3IDNX<O=~WBqL@wfyFn
zPU)7l(kttwKX8w*A3fK+Xvrl#)wh4AhjakU+#*1q|E1&m9Xvni6f%mZ<l#)FStrgr
zQaqi-Qt(c1I-3X?18|)|l14@6Zs*oYii>A!fix@PbQ&eS&cW9N_Ih_#A_btomRT+G
zUu5XiJJFPPDin7jDLMqGd}+RpN=b))&}EXKp{ztqyNzJujeQ06_3H8(W@-E(@hx|8
zVK%K8Y_5f?b*q{BlTR{BZlzSS8(M$Gzl>sE=Ln79TUa|M1B|U5wqtwjJ^bFmZ_B1u
z-oTeW<u`iXh`k43AkQK85O=`Lm?6$g%5U6!$>ugVh&#0Q0;AF%JB2&0VJ^|$!k``}
z&knjK3P2v8i*8+6{b#q88&4=1C|3$%2VVbKv7aWX3|jl2KK}GDeDKlX?T-%c|3bM@
zUYun@v1z7i8WMMIE3ZiNLIOYtZQ2<!clbQ!#YlO1I1=%2R0Li;8HRum59&lkgF?u!
zd5OM<@gy3G5NdMDqlV?1ZPqfIre{iu{__vBSuz3G<XPAs5ui?Gbar!p6i=fgQH1xC
z<WZc6{O11855sIayt)7J0g-!qd$+>99yswo;NEpLpnSegZ{&P(@fcx-)+&u5YQamC
zl$WM1B1pB)vnZJf5I(a14?hSAjHnN+7$`@BM%5mPsF+Ei>OZ9!k7<G_zXRQuciajq
z9fL|>f+;%Z8%%y5^#5(7UW`n9at2f`O_bjw3*Muqi$>KS(8vF1T_wJ>E{D*{x*q6t
z)}?bIr0T5u8xSR%JCz$t2l#VOM+Nfj0yX)IlX<PMQ9WwA233iLHTQsFjOO*iDeDWf
za3NV6d!@I41ZTfzb03`gxM~}v&n+B?{nDH_N_X5W8{>@yBEmGD-uT=Ar?oIpjW)+y
z<87k2gM)JqbcX1%^M*d6n>tKa1C2O%tG3)INN=@gonE?a?K@S^&Wcew4dYa`jvrO^
z`*Rrs%F2<^bSN}J_0dW4kaDS%M}_jDJk-9_?pL*+oJNxjHyiI!b?~`zo|7K6eS~xy
zB?!Xf^hn2)a{5V{3uWbJxpMP@r!(9+2FWZRb(>WtyG~2_r{Ms&pGHX7?}8z}Mcg{^
zme&Bif|OJvgW6O)N2Co^oe1rXph4tv->N1m6E7onqgY;7^lE3&l~V;ad>h5M#kQEo
zTz19XVQo-8lZ1)$5B|rHeV@YgeEWEyjp(hG%wPQu7l`%6#E0s2<o*%^g4~c=w$Smb
zX={P>2oMnmDN#o>1B7r<0mrS<;uwXTuBUTnPVn<kbQvJV09HV+tZ`?T^!iAw9{L72
zuI51dyiVm=Uw3LKL2rO+HP8;OAc&wuQ4K|w_D4C=!cB6}5|tW=tAUom5mVHv%}4R6
z;E%<02^c~GWnAUx1X983YQ;lN$!5CF@>w6yTtoE`IUZwM=lUyIBy$nwPy&Q3Pg|k3
z^;0fF89kRBTG;L*<9B^|4}$<0>qlACkBc*Hw{o7&#7t1#(3;IKXyeLT0hq6p$~}^;
zomQEhh{_$5HY#}O`)Hk%XjJ_XA9A{b8m!6MHX*6ST5g-Q9eD^Yyhd3Y-9zebV0~N=
z(uuxaN+fqfZlX~R9G#s>x=G8oiQyMwnx_(V>v@$7e@xu%NmOJ>S|srk+`)-b?y|SC
zqzIay9fl9@AKwpucK`U}pt<~4qKmn{DeEh!uG>2Nl^MOx02=?V@4MRk3n8Ulpm*=x
nouqu0h<miN{JUr<lvY1P7qeZj;cfUmuhFzT7r$<kZF&C!%5slT


From d1ec5494c2f933bbeeda5e63852dbee66d9342f6 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:00:52 +0200
Subject: [PATCH 06/28] Delete __init__.cpython-36.pyc

---
 Work/tfRecord/__pycache__/__init__.cpython-36.pyc | Bin 205 -> 0 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)
 delete mode 100644 Work/tfRecord/__pycache__/__init__.cpython-36.pyc

diff --git a/Work/tfRecord/__pycache__/__init__.cpython-36.pyc b/Work/tfRecord/__pycache__/__init__.cpython-36.pyc
deleted file mode 100644
index 14365e91ca043da0883b22aa9a2927f61aeaa8eb..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 205
zcmXr!<>flW_&my#fq~&M5W@izkmUfx#WFx5g&~D8has0Sijfh-W&*OAqL_ee=3oX*
zmY0k`C7O)4I6~5jQj_zGQh;p!6d)@lHNFHY<fqAei=hZ)Xc04zSjkYt3SxtZU&hH+
zF`>n&Ma5v!&?v?wwK%&ZzaS<gwYbFBC#EDV2yARje0*kJW=VX!UP0w84x8Nkl+v73
MJCNnYAcJ`r0bnUIT>t<8


From 165d4d54d0482789c04279f396039a8f86905ff4 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:01:01 +0200
Subject: [PATCH 07/28] Delete decode_tfrecord.cpython-36.pyc

---
 .../__pycache__/decode_tfrecord.cpython-36.pyc   | Bin 3051 -> 0 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)
 delete mode 100644 Work/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc

diff --git a/Work/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc b/Work/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
deleted file mode 100644
index 34952f566042af9c4999d1a4f01997fd772b8c41..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 3051
zcmbtWTW{1x6!wg->ufGSproaAd!emcSZIK%LJ+D#^Uy*>3l-85WSpJZT`#+~JD!9j
z+I<RAcxe7X|3LrJJXPvLmA5|kr9Efty-=kB30re!d^|pL=FB<Y=gYIRe(jq-8^(fR
z{9#O8CXBy^mOlW&3`TlJgXo?%Ox;@zOZRryX}EgiH7cyiEM`9^4WG?2mw7OnVRNj)
zd>B>v44c^_1gzaRg84(RY^)s;kcY6m(Zn_JvY_D@TfPWl-`FSnv}H2F=yPvq4ovvo
zTx-HBec2!>v=h+sH4u8DF*KynC0+W0?2#cEkgho}%8~VgmLvPcNaySsLpq>cch6X-
zLvv^i%z@RS#Zu@cUJf)|?mo198w4)VMoAetpyMa-^?)+^D;*NhSp#E0_UI7b+XKqX
zCH%qxjS0OA%>%8awUs@}Y~~!$p`{~e4DEqEFu`4u!Omk+I#pQ<h_ad7$#_7O9rvR)
zS9UMj;5~sJ3Uu`l!=mT$Om>>1U7!}Dt9S7>48Q|2(7S-utwn>Grwt4lY~J3zaw|=?
zxRA9cdXXo!mPq@xN0-jmmakrYT)VJRdvxJa{p$G|h{v^7FO6hgZ-Z61{<?Yi4|L_z
z?>E~E`pRwwrlPW?tY(x;<?4?vF9)>Y>NUb`y@%fnF$-V+eXV(I{imFZT+1((*KhOu
zsZ6u=2RxT|@2tz#kGz=*w$7kqJd`b=pX=F<^1~z*{iql3a@=W-*=5jEWcLR>D`b(3
z!Z^t`W!OsU*L!I*>g6l-qf}WWavqkZCo{+P17G-HT+Oe>+nn9u$tsWJrr^rSrHGR@
zoKhmcyrLEhLpi8=JmlL^KkISjm(GjmnX)%w=2yy16CQR_DQ3_teho2lR7JK*FO=6R
zmdKS;_@ydE0KtmR05K1r>L?as4lT`w(_6vf%TX9~jx5qDJx<P0&-Uo<jW-HW;1%Jp
zFvM#9!D?|xx8W)0QknfIL$5Q~fwDIu+3cuO{pcwVao0I)JB%e4kxYg1+FXVS-<E+b
z&fpGQw24-{(HyaadPf&P4%y{vi(wAut}3m#$CIefbLA(S{g7vAvy)E+q%uBk3|A2h
z;w@H*Pkm?}snXq#UoRNiYG8>^!H@`0e1hUL6qw9`Stus&^^u%7!a7bb^L3#nK8L5n
z8kB7l%NMT-G)(7P(DGkEU>SGm0PydCXs|K67NDBBoeb!0<8ccy4FJeO5S9*%-LA*z
zlF_YneFP-{8GEb;IGWOrX4{~hDP|Ad(t<OYIrM%$`w-v*fWt+Ava(2al+)pHyCVhS
zkFuY|Om-$16;z8fsbo8{lO|!Ds^jAXpQL~-MHDB>{kANZ`FmK6s{lp!i{dh&Jc=Z6
z%c4-euD7rk1IE}}QEwBfY-30AT&seL^0&iLYT&pTWl=MhJB|6Gm?o>cWtBEuC}k`@
zQHQ;Viy#_)+;4|Pg%>4<=Qf>Of=5iRvkWit(;y5ohlvIE0`({~bB;Rnv?*@DyQ7H^
z3v}OTk^dt>OoTBB;zIu+!4b_(kzh%9Hm1mkQ%4l>648C9&0%(xR7VeJ44euYFgK<)
z3rLJ;60?!on8~c?2IL%M9cC6I13T7~vZhZzAOf8pvl1@-YF07^KwEFjN&;)=C|1(B
zsPR@*VG8)D$@JwT)u8$iUQgKM-c<cIBe})W_8XFkQ@DcXE{>tV-y4DdL*gWez!8{9
zI^zX0n|KEW;@5~olOhSV?+7v)I@e&3<JJtSvEBrvw|1|;iTmIR@}77X#aR^Zq4)sB
zhadnA5wyieD9)iktcK8P44k-!7<m_3;U4i$0{+?_0U|p0TJF6Pddn3x-~z@#85i_l
z9@zJ|Hm0OOtM7{y=y%_$Uo2k&YRJ0&27Wa1fC?-&f%ys>Igj16YF7D@CwVI1&*oWy
zZ#FOw{EyHiQ#8%hv3?}8UMhR>2K=VrNI7M1W<8+C=GL;J%`J`l8W*%jQ%5=uobs;s
bQ?}XTE9g^x90WuSZl8Frf6ia>18?SETyx}{


From b3d8b1bfa49b7bcc5a0d976999a53e447809f440 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:01:16 +0200
Subject: [PATCH 08/28] Delete NN.py

---
 Work/numpyrec/NN.py | 100 --------------------------------------------
 1 file changed, 100 deletions(-)
 delete mode 100644 Work/numpyrec/NN.py

diff --git a/Work/numpyrec/NN.py b/Work/numpyrec/NN.py
deleted file mode 100644
index 302af5d..0000000
--- a/Work/numpyrec/NN.py
+++ /dev/null
@@ -1,100 +0,0 @@
-import tensorflow as tf
-import numpy as np
-T_HEIGHT = 256
-T_WIDTH = 256
-
-TRAIN_FILE = 'tftrain.tfrecords'
-TEST_FILE = 'tftest.tfrecords'
-
-learning_rate = 0.001
-num_steps = 1500
-batch_size = 16
-display_step = 10
-
-n_hidden_1 = 3 # 1st layer number of neurons
-n_hidden_2 = 3 # 2nd layer number of neurons
-num_input = 49152 #  data input (img shape: 256*256)
-num_classes = 2 # 
-
-# tf Graph input
-X = tf.placeholder("float", [None, num_input])
-Y = tf.placeholder(tf.int64, [None, num_classes])
-
-# Store layers weight & bias
-weights = {
-    'h1': tf.Variable(tf.random_normal([num_input, n_hidden_1])),
-    'h2': tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
-    'out': tf.Variable(tf.random_normal([n_hidden_2, num_classes]))
-}
-biases = {
-    'b1': tf.Variable(tf.random_normal([n_hidden_1])),
-    'b2': tf.Variable(tf.random_normal([n_hidden_2])),
-    'out': tf.Variable(tf.random_normal([num_classes]))
-}
-
-
-# Create model
-def neural_net(x):
-    # Hidden fully connected layer with 256 neurons
-    layer_1 = tf.add(tf.matmul(x, weights['h1']), biases['b1'])
-    z1 = tf.nn.relu(layer_1)
-    # Hidden fully connected layer with 256 neurons
-    layer_2 = tf.add(tf.matmul(z1, weights['h2']), biases['b2'])
-    # Output fully connected layer with a neuron for each class
-    '''z2 = tf.nn.relu(layer_2)'''
-    out_layer = tf.matmul(layer_2, weights['out']) + biases['out']
-    return out_layer
-
-# Construct model
-logits = neural_net(X)
-prediction = tf.nn.softmax(logits)
-
-# Define loss and optimizer
-loss_op = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(
-   labels=Y, logits=logits))
-optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
-train_op = optimizer.minimize(loss_op)
-
-# Evaluate model
-correct_pred = tf.equal(prediction, tf.argmax(Y, 1))
-accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
-####
-
-
-# Initialize the variables (i.e. assign their default value)
-init = tf.global_variables_initializer()
-
-# Start training
-with tf.Session() as sess:
-
-    # Run the initializer
-    sess.run(init)
-    
-    for step in range(1, num_steps+1):
-        img, label = sess.run([features['image'], labels])
-        
-        # Run optimization op (backprop)
-        sess.run(train_op, feed_dict={X: img, Y: label})
-        
-        if step % display_step == 0 or step == 1:
-            # Calculate batch loss and accuracy
-            loss, acc = sess.run([loss_op, accuracy], feed_dict={X: img,
-                                                                 Y: label})
-            print("Step " + str(step) + ", Minibatch Loss= " + \
-                  "{:.4f}".format(loss) + ", Training Accuracy= " + \
-                  "{:.3f}".format(acc))
-
-    print("Optimization Finished!")
-    print("Testing Accuracy:", \
-            sess.run(accuracy, feed_dict={X: img,
-                                      Y: label}))
-    for i in range(750):
-        img = sess.run([a['image']])
-        print("prediction :", \
-                sess.run(prediction, feed_dict={X:img}))
-    print(sess.run([weights['h1'], biases['b1'], weights['h2'], biases['b2']]))
-    print(sess.run([weights['out'],biases['out']]))
-    
-
-
-    
\ No newline at end of file

From 6e0fd71abdcce31a461942bf535b5a0bba092ea2 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:01:24 +0200
Subject: [PATCH 09/28] Delete nprecord.py

---
 Work/numpyrec/nprecord.py | 64 ---------------------------------------
 1 file changed, 64 deletions(-)
 delete mode 100644 Work/numpyrec/nprecord.py

diff --git a/Work/numpyrec/nprecord.py b/Work/numpyrec/nprecord.py
deleted file mode 100644
index fe99af1..0000000
--- a/Work/numpyrec/nprecord.py
+++ /dev/null
@@ -1,64 +0,0 @@
-import os 
-import numpy as np
-from glob import glob
-import cv2 as cv
-###############
-# Cat vs Dog #
-###############
-
-IMAGE_CHANNELS = 3
-IMAGE_HEIGHT = 128
-IMAGE_WIDTH = 128
-NUM_CLASS = 2
-gen_file_path = "train/train/"
-
-def get_paths(ur_path):
-    file_class1 = os.path.join(gen_file_path,'cat*.jpg' )
-    file_class2 = os.path.join(gen_file_path, 'dog*.jpg')
-    cat_files = glob(file_class1)
-    dog_file = glob(file_class2)
-    size = len(cat_files) + len(dog_file)
-    return cat_files, dog_file, size
-
-def get_numpy(class1_files, class2_files, size):
-    dataX = np.zeros((size,IMAGE_HEIGHT, IMAGE_WIDTH,IMAGE_CHANNELS), dtype=np.float64)
-    dataY = np.zeros((size))
-    i = 0
-    for file in class1_files:
-            img = cv.imread(file)
-            img = cv.resize(img,(IMAGE_HEIGHT, IMAGE_WIDTH))
-            img_np = np.array(img)
-            dataX[i] = img_np
-            dataY[i]= 0
-            i += 1
-    for file in class2_files:
-        try:
-            img = cv.imread(file)
-            img = cv.resize(img, (IMAGE_HEIGHT, IMAGE_WIDTH))
-            img_np = np.array(img)
-            dataX[i] = img_np
-            dataY[i] = 1
-            i +=1
-        except:
-            continue
-    return dataX, dataY
-
-def make_npfile(dir, inputs, labels):
-    ######
-    #Space dedicated for pre-processing methods
-    ######
-    if  (dir == ""):
-        dir = os.getcwd()
-    np.save(dir, [inputs, labels], ["inputs, labels"])
-    return 0
-
-def load_npfile(filename):
-    f = os.path.join(os.path.dirname(filename), filename)
-    data = np.load(f)
-    features = data[0]
-    labels = data[1]
-    return features, labels
-alpha, beta, gamma = get_paths(gen_file_path)
-a , b = get_numpy(alpha, beta, gamma)
-make_npfile("", a, b)
-'''a, b= load_npfile("train.npy")'''

From 12c046e6b6d369a939f5ef1ce053de10919fe3fb Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:01:34 +0200
Subject: [PATCH 10/28] Delete norma.py

---
 Work/utils_data/norma/norma.py | 7 -------
 1 file changed, 7 deletions(-)
 delete mode 100644 Work/utils_data/norma/norma.py

diff --git a/Work/utils_data/norma/norma.py b/Work/utils_data/norma/norma.py
deleted file mode 100644
index a0c0fb6..0000000
--- a/Work/utils_data/norma/norma.py
+++ /dev/null
@@ -1,7 +0,0 @@
-import tensorflow as tf
-
-
-def normalize(image, label):
-  """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
-  image = tf.cast(image, tf.float32) * (1. / 255) - 0.5
-  return image, label
\ No newline at end of file

From 5a2619bfc9086058527e1f637b2215831363d8f0 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:01:42 +0200
Subject: [PATCH 11/28] Delete CNN_XRAY.py

---
 Work/CNN_XRAY.py | 227 -----------------------------------------------
 1 file changed, 227 deletions(-)
 delete mode 100644 Work/CNN_XRAY.py

diff --git a/Work/CNN_XRAY.py b/Work/CNN_XRAY.py
deleted file mode 100644
index 90da2f1..0000000
--- a/Work/CNN_XRAY.py
+++ /dev/null
@@ -1,227 +0,0 @@
-import tensorflow as tf
-import tfRecord as trec
-
-
-T_HEIGHT = 32
-T_WIDTH = 32
-T_channels = 3
-
-TRAIN_FILE = 'tftrain.tfrecords'
-TEST_FILE = 'tftest.tfrecords'
-
-learning_rate = 0.001
-
-num_steps = 1500
-
-batch_size = 8
-
-display_step = 10
-
-
-
-# Network Parameters
-
-num_input = T_HEIGHT * T_WIDTH * T_channels
-
-num_classes = 2 
-
-
-
-
-# tf Graph input
-
-X = tf.placeholder(tf.float32, [None, num_input])
-
-Y = tf.placeholder(tf.int64, [None, num_classes])
-
-
-
-
-
-
-
-# Create some wrappers for simplicity
-
-def conv2d(x, W, b, strides=1):
-
-    # Conv2D wrapper, with bias and relu activation
-
-    x = tf.nn.conv2d(x, W, strides=[1, strides, strides, 1], padding='SAME')
-
-    x = tf.nn.bias_add(x, b)
-
-    return tf.nn.relu(x)
-
-
-
-
-
-def maxpool2d(x, k=2):
-
-    # MaxPool2D wrapper
-
-    return tf.nn.max_pool(x, ksize=[1, k, k, 1], strides=[1, k, k, 1],
-
-                          padding='SAME')
-
-
-
-
-
-# Create model
-
-def conv_net(x, weights, biases):
-
-
-    x = tf.reshape(x, shape=[-1, T_HEIGHT, T_WIDTH, T_channels])
-
-
-
-    # Convolution Layer
-
-    conv1 = conv2d(x, weights['wc1'], biases['bc1'])
-
-    # Max Pooling (down-sampling)
-
-    conv1 = maxpool2d(conv1, k=2)
-
-
-
-    # Convolution Layer
-
-    conv2 = conv2d(conv1, weights['wc2'], biases['bc2'])
-
-    # Max Pooling (down-sampling)
-
-    conv2 = maxpool2d(conv2, k=2)
-
-
-
-
-    # Fully connected layer
-
-    # Reshape conv2 output to fit fully connected layer input
-
-    fc1 = tf.reshape(conv2, [-1, weights['wd1'].get_shape().as_list()[0]])
-
-    fc1 = tf.add(tf.matmul(fc1, weights['wd1']), biases['bd1'])
-
-    fc1 = tf.nn.relu(fc1)
-
-    # Output, class prediction
-
-    out = tf.add(tf.matmul(fc1, weights['out']), biases['out'])
-
-    return out
-
-
-
-# Store layers weight & bias
-
-weights = {
-
-    # 5x5 conv, 1 input, 32 outputs
-
-    'wc1': tf.Variable(tf.random_normal([3, 3, 3, 32])),
-
-    # 5x5 conv, 32 inputs, 64 outputs
-
-    'wc2': tf.Variable(tf.random_normal([3, 3, 32, 64])),
-
-    # fully connected, 32*32*64 inputs, 1024 outputs
-
-    'wd1': tf.Variable(tf.random_normal([8*8*64, 512])),
-
-    # 1024 inputs, 10 outputs (class prediction)
-
-    'out': tf.Variable(tf.random_normal([512, num_classes]))
-
-}
-
-
-
-biases = {
-
-    'bc1': tf.Variable(tf.random_normal([32])),
-
-    'bc2': tf.Variable(tf.random_normal([64])),
-
-    'bd1': tf.Variable(tf.random_normal([512])),
-
-    'out': tf.Variable(tf.random_normal([num_classes]))
-
-}
-
-
-
-# Construct model
-
-logits = conv_net(X, weights, biases)
-
-prediction = tf.nn.softmax(logits)
-
-
-
-# Define loss and optimizer
-
-loss_op = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(
-                         logits=logits, labels=Y))
-
-optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
-
-train_op = optimizer.minimize(loss_op)
-
-
-
-
-
-# Evaluate model
-
-correct_pred = tf.equal(tf.argmax(prediction, 1), tf.argmax(Y, 1))
-
-accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.int64))
-
-
-
-# Initialize the variables (i.e. assign their default value)
-tfrec_dev_input_fn = trec.decode_tfrecord.tfrec_data_input_fn(["catvdog.tfrecords"],batch_size=batch_size)
-features, labels = tfrec_dev_input_fn()
-
-init = tf.global_variables_initializer()
-
-
-
-# Start training
-
-with tf.Session() as sess:
-
-
-
-    # Run the initializer
-
-    sess.run(init)
-
-
-
-    for step in range(1, num_steps+1):
-        img, label = sess.run([features['image'], labels])
-        
-        # Run optimization op (backprop)
-        sess.run(train_op, feed_dict={X: img, Y: label})
-        
-        if step % display_step == 0 or step == 1:
-            # Calculate batch loss and accuracy
-            loss, acc = sess.run([loss_op, accuracy], feed_dict={X: img,
-                                                                 Y: label})
-            print("Step " + str(step) + ", Minibatch Loss= " + \
-                  "{:.4f}".format(loss) + ", Training Accuracy= " + \
-                  "{:.3f}".format(acc))
-
-    print("Optimization Finished!")
-    '''tfrec_dev_input_fn = trec.decode_tfrecord.tfrec_data_input_fn(["catvdogtest.tfrecords"], batch_size=1000)
-    features, labels = tfrec_dev_input_fn()
-    img, label = sess.run([features['image'], labels])
-    print("Testing Accuracy:", \
-            sess.run(accuracy, feed_dict={X: img,
-                                      Y: label}))'''
-    sess.close()
\ No newline at end of file

From fb224fd623a531fbe234a6c52347e6c0028c61df Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:01:50 +0200
Subject: [PATCH 12/28] Delete NN-Xray.py

---
 Work/NN-Xray.py | 112 ------------------------------------------------
 1 file changed, 112 deletions(-)
 delete mode 100644 Work/NN-Xray.py

diff --git a/Work/NN-Xray.py b/Work/NN-Xray.py
deleted file mode 100644
index 40788bf..0000000
--- a/Work/NN-Xray.py
+++ /dev/null
@@ -1,112 +0,0 @@
-import tensorflow as tf
-import numpy as np
-import tfRecord as trec
-T_HEIGHT = 256
-T_WIDTH = 256
-
-TRAIN_FILE = 'tftrain.tfrecords'
-TEST_FILE = 'tftest.tfrecords'
-
-learning_rate = 0.001
-num_steps = 1500
-batch_size = 16
-display_step = 10
-
-n_hidden_1 = 3 # 1st layer number of neurons
-n_hidden_2 = 3 # 2nd layer number of neurons
-num_input = 3072 #  data input (img shape: 256*256)
-num_classes = 2 # 
-
-# tf Graph input
-X = tf.placeholder("float", [None, num_input])
-Y = tf.placeholder(tf.int64, [None, num_classes])
-
-# Store layers weight & bias
-weights = {
-    'h1': tf.Variable(tf.random_normal([num_input, n_hidden_1])),
-    'h2': tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
-    'out': tf.Variable(tf.random_normal([n_hidden_2, num_classes]))
-}
-biases = {
-    'b1': tf.Variable(tf.random_normal([n_hidden_1])),
-    'b2': tf.Variable(tf.random_normal([n_hidden_2])),
-    'out': tf.Variable(tf.random_normal([num_classes]))
-}
-
-
-# Create model
-def neural_net(x):
-    # Hidden fully connected layer with 256 neurons
-    layer_1 = tf.add(tf.matmul(x, weights['h1']), biases['b1'])
-    z1 = tf.nn.relu(layer_1)
-    # Hidden fully connected layer with 256 neurons
-    layer_2 = tf.add(tf.matmul(z1, weights['h2']), biases['b2'])
-    # Output fully connected layer with a neuron for each class
-    '''z2 = tf.nn.relu(layer_2)'''
-    out_layer = tf.matmul(layer_2, weights['out']) + biases['out']
-    return out_layer
-
-# Construct model
-logits = neural_net(X)
-prediction = tf.nn.softmax(logits)
-
-# Define loss and optimizer
-loss_op = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(
-   labels=Y, logits=logits))
-optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
-train_op = optimizer.minimize(loss_op)
-
-# Evaluate model
-correct_pred = tf.equal(tf.argmax(prediction,1), tf.argmax(Y, 1))
-accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
-####
-
-
-
-
-
-tfrec_dev_input_fn = trec.decode_tfrecord.tfrec_data_input_fn(["catvdog.tfrecords"])
-features, labels = tfrec_dev_input_fn()
-
-# Initialize the variables (i.e. assign their default value)
-init = tf.global_variables_initializer()
-
-# Start training
-with tf.Session() as sess:
-
-    # Run the initializer
-    sess.run(init)
-    
-    for step in range(1, num_steps+1):
-        img, label = sess.run([features['image'], labels])
-        
-        # Run optimization op (backprop)
-        sess.run(train_op, feed_dict={X: img, Y: label})
-        
-        if step % display_step == 0 or step == 1:
-            # Calculate batch loss and accuracy
-            loss, acc = sess.run([loss_op, accuracy], feed_dict={X: img,
-                                                                 Y: label})
-            print("Step " + str(step) + ", Minibatch Loss= " + \
-                  "{:.4f}".format(loss) + ", Training Accuracy= " + \
-                  "{:.3f}".format(acc))
-
-    print("Optimization Finished!")
-    tfrec_dev_input_fn = trec.decode_tfrecord.tfrec_data_input_fn(["catvdogtest.tfrecords"], batch_size=1000)
-    features, labels = tfrec_dev_input_fn()
-    img, label = sess.run([features['image'], labels])
-    print("Testing Accuracy:", \
-            sess.run(accuracy, feed_dict={X: img,
-                                      Y: label}))
-    '''tfrec_dcatvdog= trec.decode_tfrecord.tfrec_data_catvdog(["catvdogtest2.tfrecords"])
-    a = tfrec_dcatvdog()
-    for i in range(750):
-        img = sess.run([a['image']])
-        print("prediction :", \
-                sess.run(prediction, feed_dict={X:img}))
-    print(sess.run([weights['h1'], biases['b1'], weights['h2'], biases['b2']]))
-    print(sess.run([weights['out'],biases['out']]))'''
-    
-
-
-    
\ No newline at end of file

From 1bf6206a5ad7f66cf747174300b3d349474edc33 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Wed, 2 May 2018 18:03:59 +0200
Subject: [PATCH 13/28] Add files via upload

---
 NN2-Xray.py                                   | 128 +++++++++++
 NN3-Xray..py                                  | 122 ++++++++++
 tfRecord/Tfrecord.py                          | 210 ++++++++++++++++++
 tfRecord/__init__.py                          |   2 +
 tfRecord/__pycache__/Tfrecord.cpython-36.pyc  | Bin 0 -> 4269 bytes
 tfRecord/__pycache__/__init__.cpython-36.pyc  | Bin 0 -> 250 bytes
 .../decode_tfrecord.cpython-36.pyc            | Bin 0 -> 2994 bytes
 tfRecord/decode_tfrecord.py                   | 129 +++++++++++
 utils_data/colors/colors.py                   | 116 ++++++++++
 utils_data/geotransf/geotransf.py             |  81 +++++++
 utils_data/norma/norma.py                     |  22 ++
 11 files changed, 810 insertions(+)
 create mode 100644 NN2-Xray.py
 create mode 100644 NN3-Xray..py
 create mode 100644 tfRecord/Tfrecord.py
 create mode 100644 tfRecord/__init__.py
 create mode 100644 tfRecord/__pycache__/Tfrecord.cpython-36.pyc
 create mode 100644 tfRecord/__pycache__/__init__.cpython-36.pyc
 create mode 100644 tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
 create mode 100644 tfRecord/decode_tfrecord.py
 create mode 100644 utils_data/colors/colors.py
 create mode 100644 utils_data/geotransf/geotransf.py
 create mode 100644 utils_data/norma/norma.py

diff --git a/NN2-Xray.py b/NN2-Xray.py
new file mode 100644
index 0000000..c2164b1
--- /dev/null
+++ b/NN2-Xray.py
@@ -0,0 +1,128 @@
+import tensorflow as tf
+import numpy as np
+from tfRecord.decode_tfrecord import tfrec_data_input_fn, tfrec_data_catvdog
+import matplotlib.pyplot as plt
+
+T_HEIGHT = 256
+T_WIDTH = 256
+T_Channnels = 1
+TRAIN_FILE = 'tftrain.tfrecords'
+TEST_FILE = 'tftest.tfrecords'
+
+learning_rate = 0.1
+lambda_r = 0.01
+num_steps = 1500
+num_epoch = 3
+batch_size = 16
+display_step = 10
+
+n_hidden_1 = 160  # 1st layer number of neurons
+n_hidden_2 = 160 # 2nd layer number of neurons
+
+num_input = T_Channnels*T_HEIGHT*T_WIDTH #  data input (img shape: 64*64)
+num_classes = 2 # 
+
+# tf Graph input
+X = tf.placeholder("float", [None, num_input])
+Y = tf.placeholder(tf.int64, [None, num_classes])
+
+# Store layers weight & bias
+weights = {
+    'h1': tf.Variable(tf.random_normal([num_input, n_hidden_1])),
+    'h2': tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
+    'out': tf.Variable(tf.random_normal([n_hidden_2, num_classes]))
+}
+tf.add_to_collection(tf.GraphKeys.REGULARIZATION_LOSSES, weights['h1'])
+tf.add_to_collection(tf.GraphKeys.REGULARIZATION_LOSSES, weights['h2'])
+tf.add_to_collection(tf.GraphKeys.REGULARIZATION_LOSSES, weights['out'])
+biases = {
+    'b1': tf.Variable(tf.random_normal([n_hidden_1])),
+    'b2': tf.Variable(tf.random_normal([n_hidden_2])),
+    'out': tf.Variable(tf.random_normal([num_classes]))
+}
+
+
+# Create model
+def neural_net(x):
+
+    layer_1 = tf.add(tf.matmul(x, weights['h1']), biases['b1'])
+    z1 = tf.nn.tanh(layer_1)
+
+    layer_2 = tf.add(tf.matmul(z1, weights['h2']), biases['b2'])
+
+    z2 = tf.nn.relu(layer_2)
+    out_layer = tf.matmul(z2, weights['out']) + biases['out']
+    return out_layer
+
+# Construct model
+logits = neural_net(X)
+prediction = tf.nn.softmax(logits)
+
+# Define loss and optimizer
+
+loss_op = tf.nn.softmax_cross_entropy_with_logits_v2(labels=Y, logits=logits)
+regul = tf.contrib.layers.l2_regularizer(scale=0.1)
+reg_variables = tf.get_collection(tf.GraphKeys.REGULARIZATION_LOSSES)
+reg_term = tf.contrib.layers.apply_regularization(regul, reg_variables) #Regularization term
+loss_op = tf.reduce_mean(loss_op + reg_term)
+optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)
+train_op = optimizer.minimize(loss_op)
+
+# Evaluate model
+correct_pred = tf.equal(tf.argmax(prediction,1), tf.argmax(Y, 1))
+accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
+####
+
+
+
+
+
+tfrec_dev_input_fn = tfrec_data_input_fn(["tftrain.tfrecords"], num_epochs=num_epoch, shuffle=True)
+iterator = tfrec_dev_input_fn()
+
+# Initialize the variables (i.e. assign their default value)
+init = tf.global_variables_initializer()
+
+# Start training
+with tf.Session() as sess:
+    sess.run(init)
+    for ep in range(1, num_epoch+1):
+    # Run the initializer
+        sess.run(iterator.initializer)
+        features, labels = iterator.get_next()
+    
+        img, label = sess.run([features['image'], labels])
+        for step in range(1,1+num_steps):
+            '''for i in range(16):
+                print(label[i])
+                plt.imshow(img[i].reshape(256,256))
+                plt.show()'''
+            # Run optimization op (backprop)
+            sess.run(train_op, feed_dict={X: img, Y: label})
+
+            if step % display_step == 0 or step == 1:
+            # Calculate batch loss and accuracy
+                loss, acc = sess.run([loss_op, accuracy], feed_dict={X: img,
+                                                                     Y: label})
+                print("Step " + str(step) + ", Minibatch Loss= " + \
+                "{:.4f}".format(loss) + ", Training Accuracy= " + \
+                "{:.3f}".format(acc))
+        print("Optimization finished")
+    
+    
+    tfrec_dev_input_fn = tfrec_data_input_fn(["tftest.tfrecords"], batch_size=1000)
+    iterator = tfrec_dev_input_fn()
+    sess.run(iterator.initializer)
+    features, labels = iterator.get_next()
+    img, label = sess.run([features['image'], labels])
+    print("Testing Accuracy:", \
+            sess.run(accuracy, feed_dict={X: img,
+                                      Y: label}))
+    
+    
+    tfrec_dcatvdog= tfrec_data_catvdog(["tfpred.tfrecords"])
+    a = tfrec_dcatvdog()
+    for i in range(750):
+        img = sess.run([a['image']])
+        print("prediction :", \
+                sess.run(prediction, feed_dict={X:img}))
diff --git a/NN3-Xray..py b/NN3-Xray..py
new file mode 100644
index 0000000..5b80f0b
--- /dev/null
+++ b/NN3-Xray..py
@@ -0,0 +1,122 @@
+import tensorflow as tf
+import numpy as np
+from tensorflow.examples.tutorials.mnist import input_data
+mnist = input_data.read_data_sets('MNIST_data', one_hot=True)
+import matplotlib.pyplot as plt
+
+T_HEIGHT = 28
+T_WIDTH = 28
+T_Channnels = 1
+TRAIN_FILE = 'tftrain.tfrecords'
+TEST_FILE = 'tftest.tfrecords'
+
+learning_rate = 0.1
+lambda_r = 0.01
+num_steps = 1500
+num_epoch = 3
+batch_size = 16
+display_step = 10
+
+n_hidden_1 = 160  # 1st layer number of neurons
+n_hidden_2 = 160 # 2nd layer number of neurons
+
+num_input = T_Channnels*T_HEIGHT*T_WIDTH #  data input (img shape: 64*64)
+num_classes = 10 # 
+
+# tf Graph input
+X = tf.placeholder("float", [None, num_input])
+Y = tf.placeholder(tf.int64, [None, num_classes])
+
+# Store layers weight & bias
+weights = {
+    'h1': tf.Variable(tf.random_normal([num_input, n_hidden_1])),
+    'h2': tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
+    'out': tf.Variable(tf.random_normal([n_hidden_2, num_classes]))
+}
+tf.add_to_collection(tf.GraphKeys.REGULARIZATION_LOSSES, weights['h1'])
+tf.add_to_collection(tf.GraphKeys.REGULARIZATION_LOSSES, weights['h2'])
+tf.add_to_collection(tf.GraphKeys.REGULARIZATION_LOSSES, weights['out'])
+biases = {
+    'b1': tf.Variable(tf.random_normal([n_hidden_1])),
+    'b2': tf.Variable(tf.random_normal([n_hidden_2])),
+    'out': tf.Variable(tf.random_normal([num_classes]))
+}
+
+
+# Create model
+def neural_net(x):
+
+    layer_1 = tf.add(tf.matmul(x, weights['h1']), biases['b1'])
+    z1 = tf.nn.tanh(layer_1)
+
+    layer_2 = tf.add(tf.matmul(z1, weights['h2']), biases['b2'])
+
+    z2 = tf.nn.relu(layer_2)
+    out_layer = tf.matmul(z2, weights['out']) + biases['out']
+    return out_layer
+
+# Construct model
+logits = neural_net(X)
+prediction = tf.nn.softmax(logits)
+
+# Define loss and optimizer
+
+loss_op = tf.nn.softmax_cross_entropy_with_logits_v2(labels=Y, logits=logits)
+regul = tf.contrib.layers.l2_regularizer(scale=0.1)
+reg_variables = tf.get_collection(tf.GraphKeys.REGULARIZATION_LOSSES)
+reg_term = tf.contrib.layers.apply_regularization(regul, reg_variables) #Regularization term
+loss_op = tf.reduce_mean(loss_op + reg_term)
+optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)
+train_op = optimizer.minimize(loss_op)
+
+# Evaluate model
+correct_pred = tf.equal(tf.argmax(prediction,1), tf.argmax(Y, 1))
+accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
+####
+
+
+# Initialize the variables (i.e. assign their default value)
+init = tf.global_variables_initializer()
+
+# Start training
+with tf.Session() as sess:
+    sess.run(init)
+    # Run the initializer
+    '''img, label = sess.run([features['image'], labels])'''
+    batch = mnist.train.next_batch(batch_size)
+    print(batch[0][1].shape)
+    for step in range(1,1+num_steps):
+        
+        '''for i in range(16):
+            print(label[i])
+            plt.imshow(img[i].reshape(256,256))
+            plt.show()'''
+        # Run optimization op (backprop)
+        
+        '''sess.run(train_op, feed_dict={X: batch[0], Y: batch[1]})
+
+        if step % display_step == 0 or step == 1:
+        # Calculate batch loss and accuracy
+            loss, acc = sess.run([loss_op, accuracy], feed_dict={X: batch[0],
+                                                                Y: batch[1]})
+            print("Step " + str(step) + ", Minibatch Loss= " + \
+            "{:.4f}".format(loss) + ", Training Accuracy= " + \
+            "{:.3f}".format(acc))
+    print("Optimization finished")
+    
+    
+    
+    print("Testing Accuracy:", \
+            sess.run(accuracy, feed_dict={X: mnist.test.images,
+                                      Y: mnist.test.labels}))
+    
+    
+    tfrec_dcatvdog= tfrec_data_catvdog(["tfpred.tfrecords"])
+    a = tfrec_dcatvdog()
+    for i in range(750):
+        img = sess.run([a['image']])
+        print("prediction :", \
+                sess.run(prediction, feed_dict={X:img}))
+    
+    print(sess.run([weights['h1'], biases['b1'], weights['h2'], biases['b2']]))
+    print(sess.run([weights['out'],biases['out']]))'''
\ No newline at end of file
diff --git a/tfRecord/Tfrecord.py b/tfRecord/Tfrecord.py
new file mode 100644
index 0000000..fa1399e
--- /dev/null
+++ b/tfRecord/Tfrecord.py
@@ -0,0 +1,210 @@
+from random import shuffle, seed
+import os
+import glob
+
+import numpy as np
+
+import cv2
+
+import tensorflow as tf
+
+import sys
+
+
+
+
+UR_DATA_PATH = "./train/*/*.jpg"
+UR_PRED_PATH = "./test1/test1/*.jpg"
+
+UR_TEST_SIZE = 1000
+
+
+
+
+# This function returns the different paths and corresponding labels
+def shuffling_data(data_path, labels):
+    c = list(zip(data_path, labels))
+
+    shuffle(c)
+
+    #NOTE: data and labeld are tuples. Care about immutability 
+
+    data, labeled = zip(*c) #NOTE:ith element in data correpand to ith label elmt
+
+    return data, labeled
+
+def read_paths_treat(ur_path, test_size):
+    pathname = os.path.dirname(ur_path)
+    #Reading the path of each data(ex: image) and then extracting the labels from path
+    data_path = glob.glob(ur_path)
+    '''labels = [int(i.split('.')[0][-1]) for i in data_path]'''#xray
+    labels = [i.split('\\')[-1] for i in data_path]  #Cat vs Dog
+    labels_num = [0 if "cat" in i else 1 for i in labels] #Cat vs Dog
+    len_d = len(labels_num)
+    k = int(len_d/2)
+    a = int((len_d-test_size)/2)
+    b = int(len_d-(test_size/2))
+    data_path_test, labels_test = data_path[a : k], labels_num[a:k]
+    data_path_test = data_path_test + data_path[b:]
+    labels_test = labels_test + labels_num[b:]
+    data_path_train,labels_train = (data_path[0:a]), (labels_num[0:a])
+    data_path_train = data_path_train + data_path[k:b]
+    labels_train = labels_train + labels_num[k:b]
+    data_path_test, labels_test = shuffling_data(data_path_test, labels_test)
+    data_path_train, labels_train = shuffling_data(data_path_train, labels_train)
+    return data_path_test, labels_test,data_path_train, labels_train
+
+
+def split_data(training_prop,test_prop, data, labels, dev_prop=0):
+
+    ''' 
+    - training_prop is required .numerical float < 1
+    - dev_prop is optional. Numerical float < 0.3
+    - test_prop is required. Numerical float < 0.3
+    - data and labels are lists 
+    '''
+
+    #NOTE: data is a list of paths of each image
+
+    training_data = data[0:int(training_prop*len(data))]
+
+    training_labels = labels[0:int(training_prop*len(labels))]
+
+
+
+    if dev_prop != 0:
+
+        dev_data = data[int(training_prop*len(data)):int((1-dev_prop)*len(data))]
+
+        dev_labels = labels[int(training_prop*len(labels)):int((1-dev_prop)*len(labels))]
+
+        test_data = data[int((1-dev_prop)*len(data)):]
+
+        test_labels = labels[int((1-dev_prop)*len(labels)):]
+
+        return training_data,training_labels, dev_data, dev_labels, test_data, test_labels
+
+    else:
+
+        test_data = data[int((1-test_prop)*len(data)):]
+
+        test_labels = labels[int((1-test_prop)*len(labels)):]
+
+        return training_data, training_labels, test_data, test_labels
+
+
+
+# Load image 
+
+def load_image(data, s_width, s_height):
+
+    #cv2 load data from data (value of one path) 
+
+    img_load = cv2.imread(data, cv2.IMREAD_GRAYSCALE)
+   
+    '''img_after = cv2.cvtColor(img_load, cv2.COLOR_BGR2GRAY)'''
+    
+    img= cv2.resize(img_load, dsize=(s_width, s_height), interpolation=cv2.INTER_CUBIC)
+
+    img_f = img.tostring()
+    
+    return img_f
+
+
+def _int64_feature(value):
+
+    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))
+
+
+
+def _bytes_feature(value):
+
+    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))
+
+
+
+def get_tfrecord_file(ur_path, test_size, s_width, s_height, *x_filename,):
+
+    #NOTE:data is a tuple (list) of paths of each image
+
+    #NOTE: data can come from the dev-set or test-set
+
+    #NOTE: Same for labeled
+
+    test_img, test_label, train_img, train_label = read_paths_treat(ur_path, test_size)
+
+    file_path = x_filename[0]
+
+    #Open a TFRecordWriter
+
+    writer = tf.python_io.TFRecordWriter(file_path)
+    dog, cat= 0,0
+    for i in range(len(train_img)):
+        if "cat" in train_img[i]: 
+            cat +=1
+        else:
+            dog +=1
+        new_img = load_image(train_img[i], s_width, s_height)
+
+        new_label = train_label[i]
+
+        new_feature = { 'height': _int64_feature(s_height),
+                        'width' : _int64_feature(s_width),
+                        'label': _int64_feature(new_label),
+                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
+        }
+        #NOTE: PLease refer to this url for defenirtion: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
+
+        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
+
+        writer.write(example.SerializeToString())
+    print(dog,cat)
+    writer.close()
+    sys.stdout.flush()
+    file_path = x_filename[1]
+    writer = tf.python_io.TFRecordWriter(file_path)
+
+    for j in range(len(test_img)):
+
+        new_img = load_image(test_img[j], s_width, s_height)
+
+        new_label = test_label[j]
+
+        new_feature = { 'height': _int64_feature(s_height),
+                        'width' : _int64_feature(s_width),
+                        'label': _int64_feature(new_label),
+                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
+        }
+        #NOTE: PLease refer to this url for def: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
+
+        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
+
+        writer.write(example.SerializeToString())
+
+    writer.close()
+
+    sys.stdout.flush()
+
+    return 0
+
+
+def get_tfrecord_pred(ur_path, s_width, s_height, x_filename,):
+    pathname = os.path.dirname(ur_path)
+    data_path = glob.glob(ur_path)
+    file_path = x_filename
+    writer = tf.python_io.TFRecordWriter(file_path)
+    for i in range(len(data_path)):
+        new_img = load_image(data_path[i], s_width, s_height)
+        new_feature = { 'height': _int64_feature(s_height),
+                        'width' : _int64_feature(s_width),
+                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
+                        }
+        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
+        writer.write(example.SerializeToString())
+    writer.close()
+    sys.stdout.flush()
+    return 0
+
+
+'''get_tfrecord_pred(UR_PRED_PATH, 256, 256, "tfpred.tfrecords")
+''''''get_tfrecord_file(UR_DATA_PATH, UR_TEST_SIZE, 256, 256, "tftrain.tfrecords", "tftest.tfrecords")'''
\ No newline at end of file
diff --git a/tfRecord/__init__.py b/tfRecord/__init__.py
new file mode 100644
index 0000000..dab8f1b
--- /dev/null
+++ b/tfRecord/__init__.py
@@ -0,0 +1,2 @@
+from . import Tfrecord
+from . import decode_tfrecord
\ No newline at end of file
diff --git a/tfRecord/__pycache__/Tfrecord.cpython-36.pyc b/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..6446f71bb9044e55ae8e57687425f3d41c491da8
GIT binary patch
literal 4269
zcmb_fTW=gm74E9OPS3^TvEw+q3mUB!!31z_%WY-CI<d1DvL>>fjc9opb*6g8)0tbU
z+lgcK@G=4M5(H?))BXVX6^MVpbG_~h`3rf0?^Ms&&TNziM7R2M-A?tXQ{VaO^o?@a
z55IUB{N|Ejd}W-wOq4&yll%oj7=rbTJ_}f%2b@uz_sqc5Z!56$+YTK4b^}+xy`Z4q
ze&7o;C<-em2|FkYN4U?8pdvg`z<WWg2wxPRv!E(UqKuYBQ4tGx*F;q;;=LqlVhQhM
zu`Dj&eL?U&qjB*i>@pf$xyj*CyWNYFl|+%)Iaye{lFG0<xN`Z*<+XU!IaxwOl%&_D
zQa5)0jrFxSO@YhCW8t}tC%Fuf8Y#<+j7|82jd{v5ycxCeKN<&oY>utW%&a!q!FC#^
zvU=SlRpv=|Bx%lutqLMc!{#VV50%pkA4a`Iv6ix^HE?x9B>rwlAERGgYkhQZKZ#_b
z<+bYvn^E#89gYqj4CSMPw7na(hEg2tx1}zujh?AWzUc0t)1)b}c}C(v7_7oQtHfmq
z<<s?5P&r*6wh!xzS<Ev-ug;83T`~m|{JD{to78RzQ`anAvr{Xx(dwjj<_POu8sSno
zW2?rr?&Wo=tx|shKJqg6cl?r(*|QQi#{RgN`RD^WsP7_LvLZ&D)r*-appk3NO3*Bg
z%UKDUZdRCTma;N5&+6sW!Q90O$1dSJs9}VzkuH4)^|pn5b9P#AguZCY$US>_l6N3F
zMrM3o8?%hXfcgY(xn%sAKPENv6+3uIiZB0ztHG3=jC$P^IOJ<r73=C9%|#KTUis?;
zjn=R~y4j%V>06F*@_~#(q2ZA<Q<T%S_1>@*_L7?@%oN^i;Cq-s8@N1X0vLVd$YsF>
zT=1ER%4~(HBqJs5ytbyn<sy}9vxYMp<r3Nm^;6?sl8suu;VM2%l%-+pimn`leZ*O(
zH+-nfUNlf<caUnRH7d$Ik~!WB6m89<dlD&+G)YN1G^Yy@_<1-q2S@$D#w1Ol*dxV4
z#U6qbW>M3mNmOw<mMUP1vrY<dRhsouZO-wkI`82RXoc@X7_7v7USSpHu@bA|e&?K|
zc1dGf#0qrLK9c#G9aiDij0=_mBE$z1whVv}8pk$JJ8L)3wL6(5*dtT^E90+grC`y_
zTAiQA&CKbz=`82QouAjE`Q~R3rnUvKcdl=$HSO~eF%DwJ%L<v_cJ{CyJr<FCwxYRt
z^1e6CLvtW7go)s=PX1Z<>$tvO*Az_jG?K$ny_?i!^!TVNBT-+&#z(RXu-DtYVVKs}
z>({1AB6@N<emF|I!$H_vtM7bm{MEG^v(FIJ=0EXv6+%%x9Ef_3o_Z*wI)z-4n|>BI
zJBXQgvG0;7p-~LVudOM_VNTzc1sbKn?ah{OZZv%`SU6SAKO5+gy2~e>X=^^IZZ7ID
z%G>9`?Q7D})c+B_ovc6@#Z|s)dHg%L-@{$CNXl=ar$vW3T?Au-wTALl6zCuTkd2s6
zaIE2Nf?fgZHXsoLwlC<YvtoNf@V2qV99{`*p`*Ku=#S)R*b8Z^8V-2p$#vy)`xJUX
zb^FuZt&Pp*C%YTJ-n+eVcS|`kBE6uvy|cfy+q`}M*7j}XrNbnZKuW{We#Cj+eA*Q_
zH@&2J7<D^`Y2bDHohGm%KsfzQU{kgIHJ{LcX15=9q83(`ABR@OTHzj-YiK#8Q)0%`
z<B&F&yb#lh86AZkhINDUUohhP9pd;WVecqH!N-b!_+gXa(y$ay+sfAas@yx^@FN+i
z!ZwY#O9Pm^g1P6-tTgjE+PQ+n5Y4*=mp?%He>PHoj7)s^48WP1D6f$P*Gb%fP=#C6
z`Bv+4^{I8oP?}lyW0YIuHoEA@Ai#fuC;0;$N{mJM(t$O$$IjRtd&GKR#X<_#!mR~l
z1io-_E*Hn8<G-dBvITd-CT!FA{P*KBc+AbpmkgX{pJgRs{s^c2z{`rdMCZ8f&3I-`
z%y9+Wg|Xv>6kIuBkl<m+1xOFyE5v?Qp}B?ir8Tair5KmAEN4|=V{@GsY`mEHaV1-f
z7bakEW;|YhV5C*b5GF{G(6=}-kn`2zrFfZ2SdTfsp5qg}4rE4p9r3sZOYN+NjAv%4
zL$&{}rGH|Ntp{(gP!M3fot(jzu`~Mk2T+|nj+C7TmaR!eA&d|!fQ&ocZ4|v}F#-4K
z1cF54f<6#6m{ei(EIk|!n%$wQ?BB^#(+9E(B2cyr2c1aMr4Bd^Xi2zH4rM@Ls=P31
zYJ_P}DtGHR?2nMUdw241NMIvPrWW^bMua`k#r|+l^MkTSpovu3t==$+l$ktBl#`@l
zc!aig?<hHJEJ;d^<Tprsm&Etz+x~I0-9@5-G$imesS=6!XY>jxrwjtz(FuxqG51h`
z^rlI?HcR4{XV{V2x0FdKcMv@#JyoDWu2rPs)Q`%IrYmIeJQFN-qO_S#lX`99Z!u+Z
z3&QYfl*S`LsUb!2xyPztF^AE;Y_0Md^U+67iP!k+eajf>b9o=LoQ5m?4V<Phe?ui_
zJe~Gi>Ffsp2@Dc^uzl=g4kh@<aq6ZXCH@-T!PrX+fVT&D<BUKc7E)h_o{!LHX%V`m
zxJ;pr&|_bkV;?P*ctOi*<`bt16QNhk+;}l7#<gkaeF!+a@sbX><@iFpLM1`@wvFE_
z`4d<y>7bMA5Fpl{=C7O7q!D^i(h(>xl2|1{e?&R^{ftT<lh`0Zd_RXyNiLUkCgxPR
zNu@0kcSsP&$!!w9An{8QcOe?39B!vblb=%8H%XAQ<qnBok<eN9Ju2;zpdXF5qhW+h
z=^q%C5TKE;Vy`X88WL5F0ltbdAwYhHet1NCVFgDlGMiyv|Ak6A^>e?FH<XRU3%`zh
zFp}S<IwD|@3?-7Vr@5zSc|qy^ZgX>Ef1`PCWB=#6c5ioU^Q^YNwYT5g+YYvL-k%c#
z<rwomCH!gQ&MlPwb|t0L%3JRb#ZfQ1Np>fcZdb7!575=OZ83{mj#u<n{D!ySIk-2x
GrT+pAi{Z%t

literal 0
HcmV?d00001

diff --git a/tfRecord/__pycache__/__init__.cpython-36.pyc b/tfRecord/__pycache__/__init__.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..acd70d6212ae83d5b3fae5227fce220a4d6fda52
GIT binary patch
literal 250
zcmXr!<>l%-@+``ffq~&M5W@izkmUfx#WFx5g&~D8has0Sijfh-W&*OAqL_ee=3oX*
zmY0k`C7O)4I6~5jQj_zGQh;p!6d)@lHNFHY<fqAei=hZ)Xc04zSjkYt3SxtZUkS-p
zF`>n&Ma3~bsd@Qj`7ti3#n~nK1u-S5#U;KzG2!_|*+wxgi6x2g0Y#|=Mfu68#l@L<
z>G8fkx+%s6x&|h?Mh1o!V0AGiX+dB&#KgyE=4F<|$LkeT-r}&y%}*)KNwovHv>4<H
G9!3DogF@#3

literal 0
HcmV?d00001

diff --git a/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc b/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..570936efb2c0e03693e276e21fb0ad73e18a506b
GIT binary patch
literal 2994
zcmbtWOK%)S5T5RN?)aI+iEI+WB1m8%v4Rt$2!ewY;{y*X5KyFzqse%?cXu|kGeb{1
zSzEIwM{*AM2O#kW_)C34;^Z4=E>P7o`v{6qB=@RodaHY?s=KScZ*I=b`R9H;X)Z4s
z#-GMiGhzHC^n4QpGZ=}DCecsYH1*SJTKZ}CoTkfWna!LRr0KCa=CK-#YHXhQYz9U?
zBM*$m!U<S1woeGi6L4p2wJg*?v!SCy?D-OiBjbo1(YDD5qc6OnRajua+-|{?Zn%&_
zKL<VE0HGHeLqi%p(xWfQAsLc_^vuGjM%GJOjqH~ropWdmX+cZV7#21MI>z$=dOiSw
zCOar8Bgb_74`Mr}jQ&Q41awwm6y%T&ao#Q{Ggt5p2Q(%;UFaTYEv>B`Qar3<I<%O}
zycfpMF6_dD@JwbFW<(0BjX~A68${Vm?q|F~l^ylN4p(*@?($e*gaSi7!DG<tcqY58
z(JoMnhopD$77QQ)Qs`ZPcWcRD=CXlD1b*y;TldoB85gqtL_2v>Z;P~Fe{|zoeQkaH
zasB#6{n7OstLxY5ARgD-aT>~ewF6e6{dM=`A9z&%{<zy&)GfQ)Fcp<;WwpXwDp$XB
zbFD#}uJ#cO^d5dU#2mc+_wCkaJ3r-I<XV2RwzJ9er!viU9;V{yj%@$HTd80>3?7UJ
zvMu!g)ofq+L6VAo7)J*j_nGUQq8G@)_j*;pLKy~8lI_W$ovhx8(^eSg8>_EVWueG<
zP??^r9NTaB!Uq>>VJjMN_AO7gcqsP-S57WPlyu;Z68Y&ZwNx6)K^60W55j&HbLCgT
zi}1O!_h991<)#S_x~UX1Xcpgw6LM5dwyP+V*DhV;$|>VgwetADioSob0I%vZEX6!p
znhU14g0+^TFz7s4qO<fYxkNqNqX%D{>WCU{c^noVv0C_XtGuL5_?L63%zl_*)EVqR
z*}I`^b=Adw_>{xZM>4`23wPmsBFRN4Q_(QRB@Et0b7;Vu%qU;KALMY$s@9HTo`ij#
zD?i!m2Rut#-F#~MwejWR;pBKE2Jx0=iBEm#o_TQf%{H!WHf(Vf42Ta<Tmhlnyt~(K
z$6S1fD$HMq9cQK%>U%i%8aKGSO47x;_z3=;)S*&aSh3h@&@qSaLC=2yftA~%1wh;}
z(fDHaEC4QZkQ8*&c-#hbGy!w0LqZJcz}W41jIJ2HTF*xu0$j1jdO)Bl{b;of+L?0o
z(5);ula)j7=gSWPF#s!E#3U;VWmh>}9(B4>Amk|fdBkLQ0!xkAGB;}3zU-z+5T)wu
zIIG83u52m7C{gZLRgEm%#}eEE^m$NLlnFH<R?vf6Rf6($kp*!Cuwp+8<2@*?-F?Y(
zt!mViKL|z{(u8p<%)(YA_nQl41x>uWRcSU|s9&r&aUDE~RS-=->UV;&tjjvXdz;QN
z!9%9~tieRS48kDum|yT)q#lKC&Qpgjo8ohrYatqA!@I>Xc3qW6Mw?VT2k2jt5zOg^
zi^n{Nj{d`QBPN^Txsua*Hm1nP+=vxk;kK`}In1t#-RLomfHpw`=EiJh0dEn_TQ;&8
zGnoa91&pj2%VwmoBh4b)`t}<{V9;YCLf5#}|0g1&fCZ-}B7wE@IwI1!s6kfD!V-v4
z^XIEqs*&^|Oiw7}{#5<8LivoPol`Q1Q-*?QEY6_7FBpNJJmNfvh9fYObjAy$GjS0G
z!q|v6lQIc)><BU%G`C@p<JJtS0p0|7w+`-{CVU75$xghD;xdYNK&T5E!I8IeAg)O*
zc^xi;gb!rH(h-UG@Rj#bVEMw4YY?4;joA4e^fGMXod<ZfJp!zB<&DBFao3uHT#djJ
zNUl6;AO`G#@-Dvxs*C&vePgN`boO0=-;e;}SZe~)bqfOi7<F`_Ys@HL@+40Me9SyA
z@y!MTf{zfIWQwM_I@1qj7N;_fcH!FuN6M+5W)@2g_Zkkg!>OAbpNsPD^i#GM^9_te
Ra~B*6{Cv;9;;;DY{{o=%+XDaq

literal 0
HcmV?d00001

diff --git a/tfRecord/decode_tfrecord.py b/tfRecord/decode_tfrecord.py
new file mode 100644
index 0000000..d2a5546
--- /dev/null
+++ b/tfRecord/decode_tfrecord.py
@@ -0,0 +1,129 @@
+#NOTE: refer to this url: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/how_tos/reading_data/fully_connected_reader.py
+import tensorflow as tf
+import numpy as np
+import os
+import matplotlib.pyplot as plt
+
+
+
+
+
+
+
+def tfrec_data_input_fn(filenames, num_epochs=1, batch_size=16, shuffle=False):
+    
+    def _input_fn():
+        def _parse_record(tf_record):
+            features = {
+                'image': tf.FixedLenFeature([], dtype=tf.string),
+                'label': tf.FixedLenFeature([], dtype=tf.int64)
+            }
+            record = tf.parse_single_example(tf_record, features)
+
+            image_raw = tf.decode_raw(record['image'], tf.uint8)
+            label = tf.one_hot(tf.cast(record['label'], tf.int32), depth=2)
+            
+            def _normalize(image_x):
+                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+                image = tf.cast(image_x, tf.float32) * (1. / 255)-0.5
+                return image
+            
+            image_raw = _normalize(image_raw)
+            '''image_raw = tf.reshape(image_raw, (32,32,3))'''#For vizualizing 
+            return { 'image': image_raw}, label
+        
+        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
+        dataset = tf.data.TFRecordDataset(filenames)
+        dataset = dataset.map(_parse_record)
+        
+        dataset = dataset.repeat(num_epochs)
+        dataset = dataset.batch(batch_size)
+
+        iterator = dataset.make_initializable_iterator()
+        '''features, labels = iterator.get_next()'''
+
+        return iterator
+    
+    return _input_fn
+
+
+def read_file(x_filename, x_capacity=800):
+    record_iterator = tf.python_io.tf_record_iterator(path=x_filename)
+
+    example = tf.train.Example()
+    for str_rec in record_iterator:
+        example.ParseFromString(str_rec)
+        height = int(example.features.feature['height']
+                                 .int64_list
+                                 .value[0])
+
+        width = int(example.features.feature['width']
+                                .int64_list
+                                .value[0])
+    
+
+        img_string = (example.features.feature['image']
+                                  .bytes_list
+                                  .value[0])
+
+        label = (example.features.feature['label'].int64_list.value[0])
+
+
+    return 0
+
+
+def tfrec_data_catvdog(filenames, num_epochs=1, batch_size=16):
+    
+    def _input_fn():
+        def _parse_record(tf_record):
+            features = {
+                'image': tf.FixedLenFeature([], dtype=tf.string),
+            }
+            record = tf.parse_single_example(tf_record, features)
+
+            image_raw = tf.decode_raw(record['image'], tf.uint8)
+            
+            
+            def _normalize(image):
+                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+                image = tf.cast(image, tf.float32) * (1. / 255)-0.5
+                return image
+
+            image_n = _normalize(image_raw)
+            
+            
+
+            return { 'image': image_n }
+        
+        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
+        dataset = tf.data.TFRecordDataset(filenames)
+        dataset = dataset.map(_parse_record)
+
+        '''dataset = dataset.repeat(num_epochs)'''
+        '''dataset = dataset.batch(batch_size)'''
+
+        iterator = dataset.make_one_shot_iterator()
+        features= iterator.get_next()
+
+        return features
+    
+    return _input_fn
+
+
+
+
+
+'''read_file("./tftest.tfrecords")
+
+tfrec_dev_input_fn = tfrec_data_input_fn(["tftest.tfrecords"])
+features, labels = tfrec_dev_input_fn()
+
+with tf.Session() as sess:
+    for step in range(2):
+        img, label = sess.run([features['image'], labels])
+        img = np.reshape(img, (16,64,64,3))
+    
+        for i in range(16):
+            plt.imshow(img[i])
+            plt.show()'''
+        
\ No newline at end of file
diff --git a/utils_data/colors/colors.py b/utils_data/colors/colors.py
new file mode 100644
index 0000000..f690558
--- /dev/null
+++ b/utils_data/colors/colors.py
@@ -0,0 +1,116 @@
+import cv2
+
+#############
+#Color change
+#############
+#Changing color space:
+def change_color_gray(frame):
+    change = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+    return change
+
+def change_color_hsv(frame):
+    change = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
+    return change
+
+def change_color_rgb(frame):
+    change = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
+    return change
+
+#Object tracking: using a specified range of color
+#Ex: lower_blue = np.array([110,50,50]); upper_blue = np.array([130,255,255])
+#lower_col & upper_col have to be numpy arrays
+def track_color(frame, lower_col, upper_col):
+    hsv = change_color_hsv(frame)
+    mask = cv2.inRange(hsv, lower_col, upper_col)
+    img = cv2.bitwise_and(frame,frame,mask=mask)
+    return img
+
+#############
+#Thresholding
+#############
+
+#Images processed in this part must be grayscale
+#Ex: if x>value_min --> white else black
+# value_aft is given if pixel > value_thr
+
+def thresh_bin(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_BINARY)
+    return img
+
+def thresh_bin_inv(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_BINARY_INV)
+    return img
+
+def thresh_truc(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TRUNC)
+    return img
+
+def thresh_tozero(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TOZERO)
+    return img
+
+def thresh_tozero_inv(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TOZERO_INV)
+    return img
+
+#Adaptive thresholding
+#block_h: means block size height, block_w means block size width
+
+##With binary thresholding
+
+def thresh_adapgauss_bin(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_BINARY, block_h, block_w )
+
+    return img
+
+def thresh_adapgauss_bininv(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_BINARY_INV, block_h, block_w )
+
+    return img
+
+def thresh_adapgauss_tunc(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_TRUNC, block_h, block_w )
+
+    return img
+
+def thresh_adapgauss_toz(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_TOZERO, block_h, block_w )
+
+    return img
+
+def thresh_adapgauss_tozinv(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_TOZERO_INV, block_h, block_w )
+
+    return img
+
+def thresh_adapmean_bin(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
+                                cv2.THRESH_BINARY, block_h,block_w)
+    return img
+
+def thresh_adapmean_trunc(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
+                                cv2.THRESH_TRUNC, block_w, block_h)
+    return img
+
+def thresh_adapmean_toz(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
+                                cv2.THRESH_TOZERO, block_h, block_w)
+    return img
+
+def thresh_adapmean_tozinv(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
+                                cv2.THRESH_TOZERO_INV, block_h, block_w)
+    return img
+
+
+#################
+#Otsu thresholding
+##################
+
+#TODO: ret2,th2 = cv2.threshold(img,0,255,cv2.THRESH_BINARY+cv2.THRESH_OTSU)
\ No newline at end of file
diff --git a/utils_data/geotransf/geotransf.py b/utils_data/geotransf/geotransf.py
new file mode 100644
index 0000000..019618d
--- /dev/null
+++ b/utils_data/geotransf/geotransf.py
@@ -0,0 +1,81 @@
+import cv2
+import numpy as np
+import matplotlib.pyplot as plt
+#Let's consider M the translation matrix. Given a translation vector (tx,ty),
+# the matrix is defined as follows: M = [1 0 tx]
+#                                       [0 1 ty]
+
+
+##################
+# Translation
+##################
+def translate(frame, tx, ty):
+    dim = frame.shape
+    M = np.float32([[1,0,tx],[0,1,ty]])
+    img = cv2.warpAffine(frame, M, dim)
+    return img
+
+################
+# Rotation
+################
+#See reference: http://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_imgproc/py_geometric_transformations/py_geometric_transformations.html
+#Angle is in degree
+def rotate(frame, angle):
+    dim = (int(frame.shape[0]/2), int(frame.shape[1]/2))
+    M = cv2.getRotationMatrix2D(dim, angle, 1) #Last argument is for scale
+    img = cv2.warpAffine(frame, M, dim)
+    return img
+
+#######################
+# Affine transformation
+#######################
+# For this func , we need three points from the original image
+# And the corresponding points in the output image 
+#list1: list of tuples of input image: 3 tuples// Same for list2 (output image) 
+def affine(frame, list1, list2):
+    '''assert type(list1) == list and len(list1)==3'''
+    col,row,_ = frame.shape
+    sub1, sub2 = [], []
+    for i in list1:
+        sub1.append([i[0],i[1]])
+    for j in list2:
+        sub2.append([j[0], j[1]])
+    ps1 = np.float32(sub1)
+    ps2 = np.float32(sub2)
+    M = cv2.getAffineTransform(ps1,ps2)
+    img = cv2.warpAffine(frame, M, dsize=(col,row))
+    return img
+
+
+
+#######################
+# Perspective transformation
+#######################
+
+def perspec(frame, list1, list2):
+    '''assert type(list1) == list and len(list1)==4'''
+    col,row,_ = frame.shape
+    sub1, sub2 = [], []
+    for i in list1:
+        sub1.append([i[0],i[1]])
+    for j in list2:
+        sub2.append([j[0], j[1]])
+    ps1 = np.float32(sub1)
+    ps2 = np.float32(sub2)
+    M = cv2.getPerspectiveTransform(ps1,ps2)
+    img = cv2.warpPerspective(frame, M, dsize=(col,row))
+
+    return img
+
+'''a = cv2.imread("C:/Users/User12/Desktop/Work/train/train/cat.0.jpg", cv2.IMREAD_UNCHANGED)
+b = perspec(a, [(50,50),(300,50),(50,300), (300,300)], [(0,0),(250,0),(0,250),(250,250)])
+plt.imshow(b)
+plt.show()'''
+
+
+#######################
+# Morphological transformation
+#######################
+#TODO:
+
+
diff --git a/utils_data/norma/norma.py b/utils_data/norma/norma.py
new file mode 100644
index 0000000..e6cd714
--- /dev/null
+++ b/utils_data/norma/norma.py
@@ -0,0 +1,22 @@
+import tensorflow as tf
+import cv2
+
+def normalize(image, label):
+  """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+  image = tf.cast(image, tf.float32) * (1. / 255) - 0.5
+  return image, label
+
+####################
+# Image Gradient
+####################
+
+def laplacian(frame):
+    img = cv2.Laplacian(frame, cv2.CV_64F)
+    return img
+
+def sobelX(frame, kernel_size):
+    img = cv2.Sobel(frame, cv2.CV_64F, 1, 0, ksize=kernel_size)
+    return img
+
+def sobelY(frame, kernel_size):
+    img = cv2.Sobel(frame, cv2.CV_64F, 0,1,ksize=kernel_size)
\ No newline at end of file

From dbc28ec4a2638f407a44f511a6b33c3795f9bd53 Mon Sep 17 00:00:00 2001
From: Amine Hilaly <hilalyamine@gmail.com>
Date: Thu, 3 May 2018 10:18:04 +0200
Subject: [PATCH 14/28] add

---
 utils_data/histogram/__init__.py  |  0
 utils_data/histogram/histogram.py | 16 ++++++++++++++++
 2 files changed, 16 insertions(+)
 create mode 100644 utils_data/histogram/__init__.py
 create mode 100644 utils_data/histogram/histogram.py

diff --git a/utils_data/histogram/__init__.py b/utils_data/histogram/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/utils_data/histogram/histogram.py b/utils_data/histogram/histogram.py
new file mode 100644
index 0000000..89b4fa4
--- /dev/null
+++ b/utils_data/histogram/histogram.py
@@ -0,0 +1,16 @@
+from matplotlib import pyplot as plt
+import cv2
+import numpy as np
+
+def plot_histogram(frame, color='b', channels=[0], mask=None, histosize=[256], ranges=[0,256]):
+    histr = cv2.calcHist([frame], channels, mask, histosize, ranges)
+    plt.plot(histr, color=color)
+    plt.show()
+
+def plot_bgr_histogram(frame, mask=None, histosize=[256], ranges=[0, 256]):
+    color = ('b','g','r')
+    for i, col in enumerate(color):
+        histr = cv2.calcHist([frame], [i], mask, [256], ranges)
+        plt.plot(histr, color=col)
+        plt.xlim([0,256])
+    plt.show()
\ No newline at end of file

From f21acc9006f0f5f900772f095534bf97d047bb91 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:44:54 +0200
Subject: [PATCH 15/28] Delete decode_tfrecord.py

---
 tfRecord/decode_tfrecord.py | 129 ------------------------------------
 1 file changed, 129 deletions(-)
 delete mode 100644 tfRecord/decode_tfrecord.py

diff --git a/tfRecord/decode_tfrecord.py b/tfRecord/decode_tfrecord.py
deleted file mode 100644
index d2a5546..0000000
--- a/tfRecord/decode_tfrecord.py
+++ /dev/null
@@ -1,129 +0,0 @@
-#NOTE: refer to this url: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/how_tos/reading_data/fully_connected_reader.py
-import tensorflow as tf
-import numpy as np
-import os
-import matplotlib.pyplot as plt
-
-
-
-
-
-
-
-def tfrec_data_input_fn(filenames, num_epochs=1, batch_size=16, shuffle=False):
-    
-    def _input_fn():
-        def _parse_record(tf_record):
-            features = {
-                'image': tf.FixedLenFeature([], dtype=tf.string),
-                'label': tf.FixedLenFeature([], dtype=tf.int64)
-            }
-            record = tf.parse_single_example(tf_record, features)
-
-            image_raw = tf.decode_raw(record['image'], tf.uint8)
-            label = tf.one_hot(tf.cast(record['label'], tf.int32), depth=2)
-            
-            def _normalize(image_x):
-                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
-                image = tf.cast(image_x, tf.float32) * (1. / 255)-0.5
-                return image
-            
-            image_raw = _normalize(image_raw)
-            '''image_raw = tf.reshape(image_raw, (32,32,3))'''#For vizualizing 
-            return { 'image': image_raw}, label
-        
-        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
-        dataset = tf.data.TFRecordDataset(filenames)
-        dataset = dataset.map(_parse_record)
-        
-        dataset = dataset.repeat(num_epochs)
-        dataset = dataset.batch(batch_size)
-
-        iterator = dataset.make_initializable_iterator()
-        '''features, labels = iterator.get_next()'''
-
-        return iterator
-    
-    return _input_fn
-
-
-def read_file(x_filename, x_capacity=800):
-    record_iterator = tf.python_io.tf_record_iterator(path=x_filename)
-
-    example = tf.train.Example()
-    for str_rec in record_iterator:
-        example.ParseFromString(str_rec)
-        height = int(example.features.feature['height']
-                                 .int64_list
-                                 .value[0])
-
-        width = int(example.features.feature['width']
-                                .int64_list
-                                .value[0])
-    
-
-        img_string = (example.features.feature['image']
-                                  .bytes_list
-                                  .value[0])
-
-        label = (example.features.feature['label'].int64_list.value[0])
-
-
-    return 0
-
-
-def tfrec_data_catvdog(filenames, num_epochs=1, batch_size=16):
-    
-    def _input_fn():
-        def _parse_record(tf_record):
-            features = {
-                'image': tf.FixedLenFeature([], dtype=tf.string),
-            }
-            record = tf.parse_single_example(tf_record, features)
-
-            image_raw = tf.decode_raw(record['image'], tf.uint8)
-            
-            
-            def _normalize(image):
-                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
-                image = tf.cast(image, tf.float32) * (1. / 255)-0.5
-                return image
-
-            image_n = _normalize(image_raw)
-            
-            
-
-            return { 'image': image_n }
-        
-        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
-        dataset = tf.data.TFRecordDataset(filenames)
-        dataset = dataset.map(_parse_record)
-
-        '''dataset = dataset.repeat(num_epochs)'''
-        '''dataset = dataset.batch(batch_size)'''
-
-        iterator = dataset.make_one_shot_iterator()
-        features= iterator.get_next()
-
-        return features
-    
-    return _input_fn
-
-
-
-
-
-'''read_file("./tftest.tfrecords")
-
-tfrec_dev_input_fn = tfrec_data_input_fn(["tftest.tfrecords"])
-features, labels = tfrec_dev_input_fn()
-
-with tf.Session() as sess:
-    for step in range(2):
-        img, label = sess.run([features['image'], labels])
-        img = np.reshape(img, (16,64,64,3))
-    
-        for i in range(16):
-            plt.imshow(img[i])
-            plt.show()'''
-        
\ No newline at end of file

From 41b00a3fd7ccb74728413277d16d6954286fe246 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:45:05 +0200
Subject: [PATCH 16/28] Delete Tfrecord.cpython-36.pyc

---
 tfRecord/__pycache__/Tfrecord.cpython-36.pyc | Bin 4269 -> 0 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)
 delete mode 100644 tfRecord/__pycache__/Tfrecord.cpython-36.pyc

diff --git a/tfRecord/__pycache__/Tfrecord.cpython-36.pyc b/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
deleted file mode 100644
index 6446f71bb9044e55ae8e57687425f3d41c491da8..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 4269
zcmb_fTW=gm74E9OPS3^TvEw+q3mUB!!31z_%WY-CI<d1DvL>>fjc9opb*6g8)0tbU
z+lgcK@G=4M5(H?))BXVX6^MVpbG_~h`3rf0?^Ms&&TNziM7R2M-A?tXQ{VaO^o?@a
z55IUB{N|Ejd}W-wOq4&yll%oj7=rbTJ_}f%2b@uz_sqc5Z!56$+YTK4b^}+xy`Z4q
ze&7o;C<-em2|FkYN4U?8pdvg`z<WWg2wxPRv!E(UqKuYBQ4tGx*F;q;;=LqlVhQhM
zu`Dj&eL?U&qjB*i>@pf$xyj*CyWNYFl|+%)Iaye{lFG0<xN`Z*<+XU!IaxwOl%&_D
zQa5)0jrFxSO@YhCW8t}tC%Fuf8Y#<+j7|82jd{v5ycxCeKN<&oY>utW%&a!q!FC#^
zvU=SlRpv=|Bx%lutqLMc!{#VV50%pkA4a`Iv6ix^HE?x9B>rwlAERGgYkhQZKZ#_b
z<+bYvn^E#89gYqj4CSMPw7na(hEg2tx1}zujh?AWzUc0t)1)b}c}C(v7_7oQtHfmq
z<<s?5P&r*6wh!xzS<Ev-ug;83T`~m|{JD{to78RzQ`anAvr{Xx(dwjj<_POu8sSno
zW2?rr?&Wo=tx|shKJqg6cl?r(*|QQi#{RgN`RD^WsP7_LvLZ&D)r*-appk3NO3*Bg
z%UKDUZdRCTma;N5&+6sW!Q90O$1dSJs9}VzkuH4)^|pn5b9P#AguZCY$US>_l6N3F
zMrM3o8?%hXfcgY(xn%sAKPENv6+3uIiZB0ztHG3=jC$P^IOJ<r73=C9%|#KTUis?;
zjn=R~y4j%V>06F*@_~#(q2ZA<Q<T%S_1>@*_L7?@%oN^i;Cq-s8@N1X0vLVd$YsF>
zT=1ER%4~(HBqJs5ytbyn<sy}9vxYMp<r3Nm^;6?sl8suu;VM2%l%-+pimn`leZ*O(
zH+-nfUNlf<caUnRH7d$Ik~!WB6m89<dlD&+G)YN1G^Yy@_<1-q2S@$D#w1Ol*dxV4
z#U6qbW>M3mNmOw<mMUP1vrY<dRhsouZO-wkI`82RXoc@X7_7v7USSpHu@bA|e&?K|
zc1dGf#0qrLK9c#G9aiDij0=_mBE$z1whVv}8pk$JJ8L)3wL6(5*dtT^E90+grC`y_
zTAiQA&CKbz=`82QouAjE`Q~R3rnUvKcdl=$HSO~eF%DwJ%L<v_cJ{CyJr<FCwxYRt
z^1e6CLvtW7go)s=PX1Z<>$tvO*Az_jG?K$ny_?i!^!TVNBT-+&#z(RXu-DtYVVKs}
z>({1AB6@N<emF|I!$H_vtM7bm{MEG^v(FIJ=0EXv6+%%x9Ef_3o_Z*wI)z-4n|>BI
zJBXQgvG0;7p-~LVudOM_VNTzc1sbKn?ah{OZZv%`SU6SAKO5+gy2~e>X=^^IZZ7ID
z%G>9`?Q7D})c+B_ovc6@#Z|s)dHg%L-@{$CNXl=ar$vW3T?Au-wTALl6zCuTkd2s6
zaIE2Nf?fgZHXsoLwlC<YvtoNf@V2qV99{`*p`*Ku=#S)R*b8Z^8V-2p$#vy)`xJUX
zb^FuZt&Pp*C%YTJ-n+eVcS|`kBE6uvy|cfy+q`}M*7j}XrNbnZKuW{We#Cj+eA*Q_
zH@&2J7<D^`Y2bDHohGm%KsfzQU{kgIHJ{LcX15=9q83(`ABR@OTHzj-YiK#8Q)0%`
z<B&F&yb#lh86AZkhINDUUohhP9pd;WVecqH!N-b!_+gXa(y$ay+sfAas@yx^@FN+i
z!ZwY#O9Pm^g1P6-tTgjE+PQ+n5Y4*=mp?%He>PHoj7)s^48WP1D6f$P*Gb%fP=#C6
z`Bv+4^{I8oP?}lyW0YIuHoEA@Ai#fuC;0;$N{mJM(t$O$$IjRtd&GKR#X<_#!mR~l
z1io-_E*Hn8<G-dBvITd-CT!FA{P*KBc+AbpmkgX{pJgRs{s^c2z{`rdMCZ8f&3I-`
z%y9+Wg|Xv>6kIuBkl<m+1xOFyE5v?Qp}B?ir8Tair5KmAEN4|=V{@GsY`mEHaV1-f
z7bakEW;|YhV5C*b5GF{G(6=}-kn`2zrFfZ2SdTfsp5qg}4rE4p9r3sZOYN+NjAv%4
zL$&{}rGH|Ntp{(gP!M3fot(jzu`~Mk2T+|nj+C7TmaR!eA&d|!fQ&ocZ4|v}F#-4K
z1cF54f<6#6m{ei(EIk|!n%$wQ?BB^#(+9E(B2cyr2c1aMr4Bd^Xi2zH4rM@Ls=P31
zYJ_P}DtGHR?2nMUdw241NMIvPrWW^bMua`k#r|+l^MkTSpovu3t==$+l$ktBl#`@l
zc!aig?<hHJEJ;d^<Tprsm&Etz+x~I0-9@5-G$imesS=6!XY>jxrwjtz(FuxqG51h`
z^rlI?HcR4{XV{V2x0FdKcMv@#JyoDWu2rPs)Q`%IrYmIeJQFN-qO_S#lX`99Z!u+Z
z3&QYfl*S`LsUb!2xyPztF^AE;Y_0Md^U+67iP!k+eajf>b9o=LoQ5m?4V<Phe?ui_
zJe~Gi>Ffsp2@Dc^uzl=g4kh@<aq6ZXCH@-T!PrX+fVT&D<BUKc7E)h_o{!LHX%V`m
zxJ;pr&|_bkV;?P*ctOi*<`bt16QNhk+;}l7#<gkaeF!+a@sbX><@iFpLM1`@wvFE_
z`4d<y>7bMA5Fpl{=C7O7q!D^i(h(>xl2|1{e?&R^{ftT<lh`0Zd_RXyNiLUkCgxPR
zNu@0kcSsP&$!!w9An{8QcOe?39B!vblb=%8H%XAQ<qnBok<eN9Ju2;zpdXF5qhW+h
z=^q%C5TKE;Vy`X88WL5F0ltbdAwYhHet1NCVFgDlGMiyv|Ak6A^>e?FH<XRU3%`zh
zFp}S<IwD|@3?-7Vr@5zSc|qy^ZgX>Ef1`PCWB=#6c5ioU^Q^YNwYT5g+YYvL-k%c#
z<rwomCH!gQ&MlPwb|t0L%3JRb#ZfQ1Np>fcZdb7!575=OZ83{mj#u<n{D!ySIk-2x
GrT+pAi{Z%t


From ff0e54e552b7bd5f9d8bb44a6b5702d2d6daa5a8 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:45:11 +0200
Subject: [PATCH 17/28] Delete __init__.cpython-36.pyc

---
 tfRecord/__pycache__/__init__.cpython-36.pyc | Bin 250 -> 0 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)
 delete mode 100644 tfRecord/__pycache__/__init__.cpython-36.pyc

diff --git a/tfRecord/__pycache__/__init__.cpython-36.pyc b/tfRecord/__pycache__/__init__.cpython-36.pyc
deleted file mode 100644
index acd70d6212ae83d5b3fae5227fce220a4d6fda52..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 250
zcmXr!<>l%-@+``ffq~&M5W@izkmUfx#WFx5g&~D8has0Sijfh-W&*OAqL_ee=3oX*
zmY0k`C7O)4I6~5jQj_zGQh;p!6d)@lHNFHY<fqAei=hZ)Xc04zSjkYt3SxtZUkS-p
zF`>n&Ma3~bsd@Qj`7ti3#n~nK1u-S5#U;KzG2!_|*+wxgi6x2g0Y#|=Mfu68#l@L<
z>G8fkx+%s6x&|h?Mh1o!V0AGiX+dB&#KgyE=4F<|$LkeT-r}&y%}*)KNwovHv>4<H
G9!3DogF@#3


From b6e846e9b0ff96a1fe35db427ed4fa3a57ac0271 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:45:17 +0200
Subject: [PATCH 18/28] Delete decode_tfrecord.cpython-36.pyc

---
 .../__pycache__/decode_tfrecord.cpython-36.pyc   | Bin 2994 -> 0 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)
 delete mode 100644 tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc

diff --git a/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc b/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
deleted file mode 100644
index 570936efb2c0e03693e276e21fb0ad73e18a506b..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 2994
zcmbtWOK%)S5T5RN?)aI+iEI+WB1m8%v4Rt$2!ewY;{y*X5KyFzqse%?cXu|kGeb{1
zSzEIwM{*AM2O#kW_)C34;^Z4=E>P7o`v{6qB=@RodaHY?s=KScZ*I=b`R9H;X)Z4s
z#-GMiGhzHC^n4QpGZ=}DCecsYH1*SJTKZ}CoTkfWna!LRr0KCa=CK-#YHXhQYz9U?
zBM*$m!U<S1woeGi6L4p2wJg*?v!SCy?D-OiBjbo1(YDD5qc6OnRajua+-|{?Zn%&_
zKL<VE0HGHeLqi%p(xWfQAsLc_^vuGjM%GJOjqH~ropWdmX+cZV7#21MI>z$=dOiSw
zCOar8Bgb_74`Mr}jQ&Q41awwm6y%T&ao#Q{Ggt5p2Q(%;UFaTYEv>B`Qar3<I<%O}
zycfpMF6_dD@JwbFW<(0BjX~A68${Vm?q|F~l^ylN4p(*@?($e*gaSi7!DG<tcqY58
z(JoMnhopD$77QQ)Qs`ZPcWcRD=CXlD1b*y;TldoB85gqtL_2v>Z;P~Fe{|zoeQkaH
zasB#6{n7OstLxY5ARgD-aT>~ewF6e6{dM=`A9z&%{<zy&)GfQ)Fcp<;WwpXwDp$XB
zbFD#}uJ#cO^d5dU#2mc+_wCkaJ3r-I<XV2RwzJ9er!viU9;V{yj%@$HTd80>3?7UJ
zvMu!g)ofq+L6VAo7)J*j_nGUQq8G@)_j*;pLKy~8lI_W$ovhx8(^eSg8>_EVWueG<
zP??^r9NTaB!Uq>>VJjMN_AO7gcqsP-S57WPlyu;Z68Y&ZwNx6)K^60W55j&HbLCgT
zi}1O!_h991<)#S_x~UX1Xcpgw6LM5dwyP+V*DhV;$|>VgwetADioSob0I%vZEX6!p
znhU14g0+^TFz7s4qO<fYxkNqNqX%D{>WCU{c^noVv0C_XtGuL5_?L63%zl_*)EVqR
z*}I`^b=Adw_>{xZM>4`23wPmsBFRN4Q_(QRB@Et0b7;Vu%qU;KALMY$s@9HTo`ij#
zD?i!m2Rut#-F#~MwejWR;pBKE2Jx0=iBEm#o_TQf%{H!WHf(Vf42Ta<Tmhlnyt~(K
z$6S1fD$HMq9cQK%>U%i%8aKGSO47x;_z3=;)S*&aSh3h@&@qSaLC=2yftA~%1wh;}
z(fDHaEC4QZkQ8*&c-#hbGy!w0LqZJcz}W41jIJ2HTF*xu0$j1jdO)Bl{b;of+L?0o
z(5);ula)j7=gSWPF#s!E#3U;VWmh>}9(B4>Amk|fdBkLQ0!xkAGB;}3zU-z+5T)wu
zIIG83u52m7C{gZLRgEm%#}eEE^m$NLlnFH<R?vf6Rf6($kp*!Cuwp+8<2@*?-F?Y(
zt!mViKL|z{(u8p<%)(YA_nQl41x>uWRcSU|s9&r&aUDE~RS-=->UV;&tjjvXdz;QN
z!9%9~tieRS48kDum|yT)q#lKC&Qpgjo8ohrYatqA!@I>Xc3qW6Mw?VT2k2jt5zOg^
zi^n{Nj{d`QBPN^Txsua*Hm1nP+=vxk;kK`}In1t#-RLomfHpw`=EiJh0dEn_TQ;&8
zGnoa91&pj2%VwmoBh4b)`t}<{V9;YCLf5#}|0g1&fCZ-}B7wE@IwI1!s6kfD!V-v4
z^XIEqs*&^|Oiw7}{#5<8LivoPol`Q1Q-*?QEY6_7FBpNJJmNfvh9fYObjAy$GjS0G
z!q|v6lQIc)><BU%G`C@p<JJtS0p0|7w+`-{CVU75$xghD;xdYNK&T5E!I8IeAg)O*
zc^xi;gb!rH(h-UG@Rj#bVEMw4YY?4;joA4e^fGMXod<ZfJp!zB<&DBFao3uHT#djJ
zNUl6;AO`G#@-Dvxs*C&vePgN`boO0=-;e;}SZe~)bqfOi7<F`_Ys@HL@+40Me9SyA
z@y!MTf{zfIWQwM_I@1qj7N;_fcH!FuN6M+5W)@2g_Zkkg!>OAbpNsPD^i#GM^9_te
Ra~B*6{Cv;9;;;DY{{o=%+XDaq


From 4837632c1b907a3c313b5402d0840fcd4dc59fc1 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:45:27 +0200
Subject: [PATCH 19/28] Delete Tfrecord.py

---
 tfRecord/Tfrecord.py | 210 -------------------------------------------
 1 file changed, 210 deletions(-)
 delete mode 100644 tfRecord/Tfrecord.py

diff --git a/tfRecord/Tfrecord.py b/tfRecord/Tfrecord.py
deleted file mode 100644
index fa1399e..0000000
--- a/tfRecord/Tfrecord.py
+++ /dev/null
@@ -1,210 +0,0 @@
-from random import shuffle, seed
-import os
-import glob
-
-import numpy as np
-
-import cv2
-
-import tensorflow as tf
-
-import sys
-
-
-
-
-UR_DATA_PATH = "./train/*/*.jpg"
-UR_PRED_PATH = "./test1/test1/*.jpg"
-
-UR_TEST_SIZE = 1000
-
-
-
-
-# This function returns the different paths and corresponding labels
-def shuffling_data(data_path, labels):
-    c = list(zip(data_path, labels))
-
-    shuffle(c)
-
-    #NOTE: data and labeld are tuples. Care about immutability 
-
-    data, labeled = zip(*c) #NOTE:ith element in data correpand to ith label elmt
-
-    return data, labeled
-
-def read_paths_treat(ur_path, test_size):
-    pathname = os.path.dirname(ur_path)
-    #Reading the path of each data(ex: image) and then extracting the labels from path
-    data_path = glob.glob(ur_path)
-    '''labels = [int(i.split('.')[0][-1]) for i in data_path]'''#xray
-    labels = [i.split('\\')[-1] for i in data_path]  #Cat vs Dog
-    labels_num = [0 if "cat" in i else 1 for i in labels] #Cat vs Dog
-    len_d = len(labels_num)
-    k = int(len_d/2)
-    a = int((len_d-test_size)/2)
-    b = int(len_d-(test_size/2))
-    data_path_test, labels_test = data_path[a : k], labels_num[a:k]
-    data_path_test = data_path_test + data_path[b:]
-    labels_test = labels_test + labels_num[b:]
-    data_path_train,labels_train = (data_path[0:a]), (labels_num[0:a])
-    data_path_train = data_path_train + data_path[k:b]
-    labels_train = labels_train + labels_num[k:b]
-    data_path_test, labels_test = shuffling_data(data_path_test, labels_test)
-    data_path_train, labels_train = shuffling_data(data_path_train, labels_train)
-    return data_path_test, labels_test,data_path_train, labels_train
-
-
-def split_data(training_prop,test_prop, data, labels, dev_prop=0):
-
-    ''' 
-    - training_prop is required .numerical float < 1
-    - dev_prop is optional. Numerical float < 0.3
-    - test_prop is required. Numerical float < 0.3
-    - data and labels are lists 
-    '''
-
-    #NOTE: data is a list of paths of each image
-
-    training_data = data[0:int(training_prop*len(data))]
-
-    training_labels = labels[0:int(training_prop*len(labels))]
-
-
-
-    if dev_prop != 0:
-
-        dev_data = data[int(training_prop*len(data)):int((1-dev_prop)*len(data))]
-
-        dev_labels = labels[int(training_prop*len(labels)):int((1-dev_prop)*len(labels))]
-
-        test_data = data[int((1-dev_prop)*len(data)):]
-
-        test_labels = labels[int((1-dev_prop)*len(labels)):]
-
-        return training_data,training_labels, dev_data, dev_labels, test_data, test_labels
-
-    else:
-
-        test_data = data[int((1-test_prop)*len(data)):]
-
-        test_labels = labels[int((1-test_prop)*len(labels)):]
-
-        return training_data, training_labels, test_data, test_labels
-
-
-
-# Load image 
-
-def load_image(data, s_width, s_height):
-
-    #cv2 load data from data (value of one path) 
-
-    img_load = cv2.imread(data, cv2.IMREAD_GRAYSCALE)
-   
-    '''img_after = cv2.cvtColor(img_load, cv2.COLOR_BGR2GRAY)'''
-    
-    img= cv2.resize(img_load, dsize=(s_width, s_height), interpolation=cv2.INTER_CUBIC)
-
-    img_f = img.tostring()
-    
-    return img_f
-
-
-def _int64_feature(value):
-
-    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))
-
-
-
-def _bytes_feature(value):
-
-    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))
-
-
-
-def get_tfrecord_file(ur_path, test_size, s_width, s_height, *x_filename,):
-
-    #NOTE:data is a tuple (list) of paths of each image
-
-    #NOTE: data can come from the dev-set or test-set
-
-    #NOTE: Same for labeled
-
-    test_img, test_label, train_img, train_label = read_paths_treat(ur_path, test_size)
-
-    file_path = x_filename[0]
-
-    #Open a TFRecordWriter
-
-    writer = tf.python_io.TFRecordWriter(file_path)
-    dog, cat= 0,0
-    for i in range(len(train_img)):
-        if "cat" in train_img[i]: 
-            cat +=1
-        else:
-            dog +=1
-        new_img = load_image(train_img[i], s_width, s_height)
-
-        new_label = train_label[i]
-
-        new_feature = { 'height': _int64_feature(s_height),
-                        'width' : _int64_feature(s_width),
-                        'label': _int64_feature(new_label),
-                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
-        }
-        #NOTE: PLease refer to this url for defenirtion: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
-
-        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
-
-        writer.write(example.SerializeToString())
-    print(dog,cat)
-    writer.close()
-    sys.stdout.flush()
-    file_path = x_filename[1]
-    writer = tf.python_io.TFRecordWriter(file_path)
-
-    for j in range(len(test_img)):
-
-        new_img = load_image(test_img[j], s_width, s_height)
-
-        new_label = test_label[j]
-
-        new_feature = { 'height': _int64_feature(s_height),
-                        'width' : _int64_feature(s_width),
-                        'label': _int64_feature(new_label),
-                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
-        }
-        #NOTE: PLease refer to this url for def: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
-
-        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
-
-        writer.write(example.SerializeToString())
-
-    writer.close()
-
-    sys.stdout.flush()
-
-    return 0
-
-
-def get_tfrecord_pred(ur_path, s_width, s_height, x_filename,):
-    pathname = os.path.dirname(ur_path)
-    data_path = glob.glob(ur_path)
-    file_path = x_filename
-    writer = tf.python_io.TFRecordWriter(file_path)
-    for i in range(len(data_path)):
-        new_img = load_image(data_path[i], s_width, s_height)
-        new_feature = { 'height': _int64_feature(s_height),
-                        'width' : _int64_feature(s_width),
-                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
-                        }
-        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
-        writer.write(example.SerializeToString())
-    writer.close()
-    sys.stdout.flush()
-    return 0
-
-
-'''get_tfrecord_pred(UR_PRED_PATH, 256, 256, "tfpred.tfrecords")
-''''''get_tfrecord_file(UR_DATA_PATH, UR_TEST_SIZE, 256, 256, "tftrain.tfrecords", "tftest.tfrecords")'''
\ No newline at end of file

From 38c3238776ff139bfb878e82a060538d8b5556fe Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:45:33 +0200
Subject: [PATCH 20/28] Delete __init__.py

---
 tfRecord/__init__.py | 2 --
 1 file changed, 2 deletions(-)
 delete mode 100644 tfRecord/__init__.py

diff --git a/tfRecord/__init__.py b/tfRecord/__init__.py
deleted file mode 100644
index dab8f1b..0000000
--- a/tfRecord/__init__.py
+++ /dev/null
@@ -1,2 +0,0 @@
-from . import Tfrecord
-from . import decode_tfrecord
\ No newline at end of file

From 494b76dbbf1e853e9ea66a6a7d02509d3d35712f Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:45:44 +0200
Subject: [PATCH 21/28] Delete colors.py

---
 utils_data/colors/colors.py | 116 ------------------------------------
 1 file changed, 116 deletions(-)
 delete mode 100644 utils_data/colors/colors.py

diff --git a/utils_data/colors/colors.py b/utils_data/colors/colors.py
deleted file mode 100644
index f690558..0000000
--- a/utils_data/colors/colors.py
+++ /dev/null
@@ -1,116 +0,0 @@
-import cv2
-
-#############
-#Color change
-#############
-#Changing color space:
-def change_color_gray(frame):
-    change = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
-    return change
-
-def change_color_hsv(frame):
-    change = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
-    return change
-
-def change_color_rgb(frame):
-    change = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
-    return change
-
-#Object tracking: using a specified range of color
-#Ex: lower_blue = np.array([110,50,50]); upper_blue = np.array([130,255,255])
-#lower_col & upper_col have to be numpy arrays
-def track_color(frame, lower_col, upper_col):
-    hsv = change_color_hsv(frame)
-    mask = cv2.inRange(hsv, lower_col, upper_col)
-    img = cv2.bitwise_and(frame,frame,mask=mask)
-    return img
-
-#############
-#Thresholding
-#############
-
-#Images processed in this part must be grayscale
-#Ex: if x>value_min --> white else black
-# value_aft is given if pixel > value_thr
-
-def thresh_bin(frame, value_thr, value_aft):
-    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_BINARY)
-    return img
-
-def thresh_bin_inv(frame, value_thr, value_aft):
-    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_BINARY_INV)
-    return img
-
-def thresh_truc(frame, value_thr, value_aft):
-    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TRUNC)
-    return img
-
-def thresh_tozero(frame, value_thr, value_aft):
-    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TOZERO)
-    return img
-
-def thresh_tozero_inv(frame, value_thr, value_aft):
-    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TOZERO_INV)
-    return img
-
-#Adaptive thresholding
-#block_h: means block size height, block_w means block size width
-
-##With binary thresholding
-
-def thresh_adapgauss_bin(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
-                                cv2.THRESH_BINARY, block_h, block_w )
-
-    return img
-
-def thresh_adapgauss_bininv(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
-                                cv2.THRESH_BINARY_INV, block_h, block_w )
-
-    return img
-
-def thresh_adapgauss_tunc(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
-                                cv2.THRESH_TRUNC, block_h, block_w )
-
-    return img
-
-def thresh_adapgauss_toz(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
-                                cv2.THRESH_TOZERO, block_h, block_w )
-
-    return img
-
-def thresh_adapgauss_tozinv(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
-                                cv2.THRESH_TOZERO_INV, block_h, block_w )
-
-    return img
-
-def thresh_adapmean_bin(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
-                                cv2.THRESH_BINARY, block_h,block_w)
-    return img
-
-def thresh_adapmean_trunc(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
-                                cv2.THRESH_TRUNC, block_w, block_h)
-    return img
-
-def thresh_adapmean_toz(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
-                                cv2.THRESH_TOZERO, block_h, block_w)
-    return img
-
-def thresh_adapmean_tozinv(frame, value_aft, block_h, block_w):
-    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
-                                cv2.THRESH_TOZERO_INV, block_h, block_w)
-    return img
-
-
-#################
-#Otsu thresholding
-##################
-
-#TODO: ret2,th2 = cv2.threshold(img,0,255,cv2.THRESH_BINARY+cv2.THRESH_OTSU)
\ No newline at end of file

From 6374e840072eb023982b22cbfb38c33608c8aab1 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:45:53 +0200
Subject: [PATCH 22/28] Delete geotransf.py

---
 utils_data/geotransf/geotransf.py | 81 -------------------------------
 1 file changed, 81 deletions(-)
 delete mode 100644 utils_data/geotransf/geotransf.py

diff --git a/utils_data/geotransf/geotransf.py b/utils_data/geotransf/geotransf.py
deleted file mode 100644
index 019618d..0000000
--- a/utils_data/geotransf/geotransf.py
+++ /dev/null
@@ -1,81 +0,0 @@
-import cv2
-import numpy as np
-import matplotlib.pyplot as plt
-#Let's consider M the translation matrix. Given a translation vector (tx,ty),
-# the matrix is defined as follows: M = [1 0 tx]
-#                                       [0 1 ty]
-
-
-##################
-# Translation
-##################
-def translate(frame, tx, ty):
-    dim = frame.shape
-    M = np.float32([[1,0,tx],[0,1,ty]])
-    img = cv2.warpAffine(frame, M, dim)
-    return img
-
-################
-# Rotation
-################
-#See reference: http://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_imgproc/py_geometric_transformations/py_geometric_transformations.html
-#Angle is in degree
-def rotate(frame, angle):
-    dim = (int(frame.shape[0]/2), int(frame.shape[1]/2))
-    M = cv2.getRotationMatrix2D(dim, angle, 1) #Last argument is for scale
-    img = cv2.warpAffine(frame, M, dim)
-    return img
-
-#######################
-# Affine transformation
-#######################
-# For this func , we need three points from the original image
-# And the corresponding points in the output image 
-#list1: list of tuples of input image: 3 tuples// Same for list2 (output image) 
-def affine(frame, list1, list2):
-    '''assert type(list1) == list and len(list1)==3'''
-    col,row,_ = frame.shape
-    sub1, sub2 = [], []
-    for i in list1:
-        sub1.append([i[0],i[1]])
-    for j in list2:
-        sub2.append([j[0], j[1]])
-    ps1 = np.float32(sub1)
-    ps2 = np.float32(sub2)
-    M = cv2.getAffineTransform(ps1,ps2)
-    img = cv2.warpAffine(frame, M, dsize=(col,row))
-    return img
-
-
-
-#######################
-# Perspective transformation
-#######################
-
-def perspec(frame, list1, list2):
-    '''assert type(list1) == list and len(list1)==4'''
-    col,row,_ = frame.shape
-    sub1, sub2 = [], []
-    for i in list1:
-        sub1.append([i[0],i[1]])
-    for j in list2:
-        sub2.append([j[0], j[1]])
-    ps1 = np.float32(sub1)
-    ps2 = np.float32(sub2)
-    M = cv2.getPerspectiveTransform(ps1,ps2)
-    img = cv2.warpPerspective(frame, M, dsize=(col,row))
-
-    return img
-
-'''a = cv2.imread("C:/Users/User12/Desktop/Work/train/train/cat.0.jpg", cv2.IMREAD_UNCHANGED)
-b = perspec(a, [(50,50),(300,50),(50,300), (300,300)], [(0,0),(250,0),(0,250),(250,250)])
-plt.imshow(b)
-plt.show()'''
-
-
-#######################
-# Morphological transformation
-#######################
-#TODO:
-
-

From 7d52cf82b4eae288e56318f9e804a05bf20d9d5a Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:46:24 +0200
Subject: [PATCH 23/28] Delete norma.py

---
 utils_data/norma/norma.py | 22 ----------------------
 1 file changed, 22 deletions(-)
 delete mode 100644 utils_data/norma/norma.py

diff --git a/utils_data/norma/norma.py b/utils_data/norma/norma.py
deleted file mode 100644
index e6cd714..0000000
--- a/utils_data/norma/norma.py
+++ /dev/null
@@ -1,22 +0,0 @@
-import tensorflow as tf
-import cv2
-
-def normalize(image, label):
-  """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
-  image = tf.cast(image, tf.float32) * (1. / 255) - 0.5
-  return image, label
-
-####################
-# Image Gradient
-####################
-
-def laplacian(frame):
-    img = cv2.Laplacian(frame, cv2.CV_64F)
-    return img
-
-def sobelX(frame, kernel_size):
-    img = cv2.Sobel(frame, cv2.CV_64F, 1, 0, ksize=kernel_size)
-    return img
-
-def sobelY(frame, kernel_size):
-    img = cv2.Sobel(frame, cv2.CV_64F, 0,1,ksize=kernel_size)
\ No newline at end of file

From 52b8a24bb51c4089706d1b312f89c8aa34101ae8 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:50:15 +0200
Subject: [PATCH 24/28] Add files via upload

---
 image_pro/__init__.py                         |   1 +
 image_pro/parameters.py                       |   7 +
 image_pro/setup.py                            |  28 +++
 image_pro/tfRecord/Tfrecord.py                | 213 ++++++++++++++++++
 image_pro/tfRecord/__init__.py                |   1 +
 .../__pycache__/Tfrecord.cpython-36.pyc       | Bin 0 -> 4269 bytes
 .../__pycache__/__init__.cpython-36.pyc       | Bin 0 -> 250 bytes
 .../decode_tfrecord.cpython-36.pyc            | Bin 0 -> 2994 bytes
 image_pro/tfRecord/decode.py                  | 129 +++++++++++
 image_pro/utils_data/__init__.py              |   1 +
 image_pro/utils_data/colors/__init__.py       |   1 +
 image_pro/utils_data/colors/colors.py         | 117 ++++++++++
 image_pro/utils_data/contours/__init__.py     |   1 +
 image_pro/utils_data/contours/contours.py     | 128 +++++++++++
 image_pro/utils_data/geotransf/__init__.py    |   1 +
 image_pro/utils_data/geotransf/geotransf.py   |  81 +++++++
 image_pro/utils_data/norma/__init__.py        |   1 +
 image_pro/utils_data/norma/norma.py           |  22 ++
 image_pro/utils_data/zooming/__init__.py      |   1 +
 image_pro/utils_data/zooming/zoom.py          |   1 +
 20 files changed, 734 insertions(+)
 create mode 100644 image_pro/__init__.py
 create mode 100644 image_pro/parameters.py
 create mode 100644 image_pro/setup.py
 create mode 100644 image_pro/tfRecord/Tfrecord.py
 create mode 100644 image_pro/tfRecord/__init__.py
 create mode 100644 image_pro/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
 create mode 100644 image_pro/tfRecord/__pycache__/__init__.cpython-36.pyc
 create mode 100644 image_pro/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
 create mode 100644 image_pro/tfRecord/decode.py
 create mode 100644 image_pro/utils_data/__init__.py
 create mode 100644 image_pro/utils_data/colors/__init__.py
 create mode 100644 image_pro/utils_data/colors/colors.py
 create mode 100644 image_pro/utils_data/contours/__init__.py
 create mode 100644 image_pro/utils_data/contours/contours.py
 create mode 100644 image_pro/utils_data/geotransf/__init__.py
 create mode 100644 image_pro/utils_data/geotransf/geotransf.py
 create mode 100644 image_pro/utils_data/norma/__init__.py
 create mode 100644 image_pro/utils_data/norma/norma.py
 create mode 100644 image_pro/utils_data/zooming/__init__.py
 create mode 100644 image_pro/utils_data/zooming/zoom.py

diff --git a/image_pro/__init__.py b/image_pro/__init__.py
new file mode 100644
index 0000000..ab28e3f
--- /dev/null
+++ b/image_pro/__init__.py
@@ -0,0 +1 @@
+#NOTE
\ No newline at end of file
diff --git a/image_pro/parameters.py b/image_pro/parameters.py
new file mode 100644
index 0000000..e61d3f4
--- /dev/null
+++ b/image_pro/parameters.py
@@ -0,0 +1,7 @@
+UR_DATA_PATH = "./train/*/*.jpg"
+UR_PRED_PATH = "./test1/test1/*.jpg"
+UR_OPTION = "gray"
+UR_TEST_SIZE = 1000
+T_HEIGHT = 256
+T_WIDTH = 256
+T_Channnels = 1
\ No newline at end of file
diff --git a/image_pro/setup.py b/image_pro/setup.py
new file mode 100644
index 0000000..3572ce0
--- /dev/null
+++ b/image_pro/setup.py
@@ -0,0 +1,28 @@
+from setuptools import setup, find_packages
+
+
+setup(
+name='image_pro',
+version='0.0.1',
+description='Image processing and save into Tfrecords',
+author='',
+author_email='hilalyamine@gmail.com/alaa.el.bouchti@gmail.com',
+license='Open',
+packages=find_packages(include=['image_pro','image_pro.*','tfRecord', 'utils_data']),
+packages_dir={'tfrecord':'tfRecord',
+              'tfrecord.tfrecord':'tfRecord/tfrecord.py',
+              'tfrecord.decode':'tfRecord/decode.py',
+              'utils': 'utils_data',
+              'utils.colors': 'utils_data/colors',
+              'utils.colors.colors': 'utils_data/colors/colors.py',
+              'utils.contours':'utils_data/contours',
+              'utils.contours.contours':'utils_data/contours/contours.py',
+              'utils.geotransf':'utils_data/geotransf',
+              'utils.geotransf.geotransf':'utils_data/geotransf/geotransf.py',
+              'utils.norma':'utils_data/norma',
+              'utils.norma.norma':'utils_data/contours/norma.py',
+              'utils.visualize':'utils_data/contours',
+              'utils.visualize.visu':'utils_data/visualize/visu.py'            
+},
+scripts = ['parameters.py']
+)
\ No newline at end of file
diff --git a/image_pro/tfRecord/Tfrecord.py b/image_pro/tfRecord/Tfrecord.py
new file mode 100644
index 0000000..718719b
--- /dev/null
+++ b/image_pro/tfRecord/Tfrecord.py
@@ -0,0 +1,213 @@
+from random import shuffle, seed
+import os
+import glob
+from .utils_data.contours.contours import draw_all_contours
+import numpy as np
+
+import cv2
+
+import tensorflow as tf
+
+import sys
+
+
+
+
+
+
+
+
+
+# This function returns the different paths and corresponding labels
+def shuffling_data(data_path, labels):
+    c = list(zip(data_path, labels))
+
+    shuffle(c)
+
+    #NOTE: data and labeld are tuples. Care about immutability 
+
+    data, labeled = zip(*c) #NOTE:ith element in data correpand to ith label elmt
+
+    return data, labeled
+
+def read_paths_treat(ur_path, test_size):
+    pathname = os.path.dirname(ur_path)
+    #Reading the path of each data(ex: image) and then extracting the labels from path
+    data_path = glob.glob(ur_path)
+    '''labels = [int(i.split('.')[0][-1]) for i in data_path]'''#xray
+    #NOTE: labels and labels_num: modify split method to get labels// conditions "cat" or "dog"
+    labels = [i.split('\\')[-1] for i in data_path]  #Cat vs Dog
+    labels_num = [0 if "cat" in i else 1 for i in labels] #Cat vs Dog
+    len_d = len(labels_num)
+    k = int(len_d/2)
+    a = int((len_d-test_size)/2)
+    b = int(len_d-(test_size/2))
+    data_path_test, labels_test = data_path[a : k], labels_num[a:k]
+    data_path_test = data_path_test + data_path[b:]
+    labels_test = labels_test + labels_num[b:]
+    data_path_train,labels_train = (data_path[0:a]), (labels_num[0:a])
+    data_path_train = data_path_train + data_path[k:b]
+    labels_train = labels_train + labels_num[k:b]
+    data_path_test, labels_test = shuffling_data(data_path_test, labels_test)
+    data_path_train, labels_train = shuffling_data(data_path_train, labels_train)
+    return data_path_test, labels_test,data_path_train, labels_train
+
+
+def split_data(training_prop,test_prop, data, labels, dev_prop=0):
+
+    ''' 
+    - training_prop is required .numerical float < 1
+    - dev_prop is optional. Numerical float < 0.3
+    - test_prop is required. Numerical float < 0.3
+    - data and labels are lists 
+    '''
+
+    #NOTE: data is a list of paths of each image
+
+    training_data = data[0:int(training_prop*len(data))]
+
+    training_labels = labels[0:int(training_prop*len(labels))]
+
+
+
+    if dev_prop != 0:
+
+        dev_data = data[int(training_prop*len(data)):int((1-dev_prop)*len(data))]
+
+        dev_labels = labels[int(training_prop*len(labels)):int((1-dev_prop)*len(labels))]
+
+        test_data = data[int((1-dev_prop)*len(data)):]
+
+        test_labels = labels[int((1-dev_prop)*len(labels)):]
+
+        return training_data,training_labels, dev_data, dev_labels, test_data, test_labels
+
+    else:
+
+        test_data = data[int((1-test_prop)*len(data)):]
+
+        test_labels = labels[int((1-test_prop)*len(labels)):]
+
+        return training_data, training_labels, test_data, test_labels
+
+def load(ur_path, option):
+
+    if option =="gray":
+        img = cv2.imread(ur_path,cv2.IMREAD_GRAYSCALE)
+    elif option =="unchanged":
+        img = cv2.imread(ur_path, cv2.IMREAD_UNCHANGED)
+    else:
+        img = cv2.imread(ur_path,cv2.IMREAD_COLOR)
+    return img
+
+
+# Load image 
+
+def load_image(data, s_width, s_height, option):
+
+    #cv2 load data from data (value of one path) 
+
+    img_load = load(data, UR_OPTION)
+
+   
+    '''img_after = cv2.cvtColor(img_load, cv2.COLOR_BGR2GRAY)'''
+    
+    img= cv2.resize(img_cont, dsize=(s_width, s_height), interpolation=cv2.INTER_CUBIC)
+
+    img_f = img.tostring()
+    
+    return img_f
+
+
+def _int64_feature(value):
+
+    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))
+
+
+
+def _bytes_feature(value):
+
+    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))
+
+
+
+def get_tfrecord_file(ur_path, test_size, s_width, s_height,option, *x_filename,):
+
+    #NOTE:data is a tuple (list) of paths of each image
+
+    #NOTE: data can come from the dev-set or test-set
+
+    #NOTE: Same for labeled
+
+    test_img, test_label, train_img, train_label = read_paths_treat(ur_path, test_size)
+
+    file_path = x_filename[0]
+
+    #Open a TFRecordWriter
+
+    writer = tf.python_io.TFRecordWriter(file_path)
+    for i in range(len(train_img)):
+        new_img = load_image(train_img[i], s_width, s_height,option)
+
+        new_label = train_label[i]
+
+        new_feature = { 'height': _int64_feature(s_height),
+                        'width' : _int64_feature(s_width),
+                        'label': _int64_feature(new_label),
+                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
+        }
+        #NOTE: PLease refer to this url for defenirtion: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
+
+        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
+
+        writer.write(example.SerializeToString())
+
+    writer.close()
+    sys.stdout.flush()
+    file_path = x_filename[1]
+    writer = tf.python_io.TFRecordWriter(file_path)
+
+    for j in range(len(test_img)):
+
+        new_img = load_image(test_img[j], s_width, s_height,option)
+
+        new_label = test_label[j]
+
+        new_feature = { 'height': _int64_feature(s_height),
+                        'width' : _int64_feature(s_width),
+                        'label': _int64_feature(new_label),
+                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
+        }
+        #NOTE: PLease refer to this url for def: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
+
+        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
+
+        writer.write(example.SerializeToString())
+
+    writer.close()
+
+    sys.stdout.flush()
+
+    return 0
+
+
+def get_tfrecord_pred(ur_path, s_width, s_height,option, x_filename):
+    pathname = os.path.dirname(ur_path)
+    data_path = glob.glob(ur_path)
+    file_path = x_filename
+    writer = tf.python_io.TFRecordWriter(file_path)
+    for i in range(len(data_path)):
+        new_img = load_image(data_path[i], s_width, s_height,option)
+        new_feature = { 'height': _int64_feature(s_height),
+                        'width' : _int64_feature(s_width),
+                        'image': _bytes_feature(tf.compat.as_bytes(new_img))
+                        }
+        example = tf.train.Example(features=tf.train.Features(feature=new_feature))
+        writer.write(example.SerializeToString())
+    writer.close()
+    sys.stdout.flush()
+    return 0
+
+
+'''get_tfrecord_pred(UR_PRED_PATH, 256, 256,UR_OPTION, "tfpred.tfrecords")'''
+get_tfrecord_file(UR_DATA_PATH, UR_TEST_SIZE, 128, 128,UR_OPTION, "tftrain.tfrecords", "tftest.tfrecords")
\ No newline at end of file
diff --git a/image_pro/tfRecord/__init__.py b/image_pro/tfRecord/__init__.py
new file mode 100644
index 0000000..6f492a6
--- /dev/null
+++ b/image_pro/tfRecord/__init__.py
@@ -0,0 +1 @@
+#NOTE
diff --git a/image_pro/tfRecord/__pycache__/Tfrecord.cpython-36.pyc b/image_pro/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..6446f71bb9044e55ae8e57687425f3d41c491da8
GIT binary patch
literal 4269
zcmb_fTW=gm74E9OPS3^TvEw+q3mUB!!31z_%WY-CI<d1DvL>>fjc9opb*6g8)0tbU
z+lgcK@G=4M5(H?))BXVX6^MVpbG_~h`3rf0?^Ms&&TNziM7R2M-A?tXQ{VaO^o?@a
z55IUB{N|Ejd}W-wOq4&yll%oj7=rbTJ_}f%2b@uz_sqc5Z!56$+YTK4b^}+xy`Z4q
ze&7o;C<-em2|FkYN4U?8pdvg`z<WWg2wxPRv!E(UqKuYBQ4tGx*F;q;;=LqlVhQhM
zu`Dj&eL?U&qjB*i>@pf$xyj*CyWNYFl|+%)Iaye{lFG0<xN`Z*<+XU!IaxwOl%&_D
zQa5)0jrFxSO@YhCW8t}tC%Fuf8Y#<+j7|82jd{v5ycxCeKN<&oY>utW%&a!q!FC#^
zvU=SlRpv=|Bx%lutqLMc!{#VV50%pkA4a`Iv6ix^HE?x9B>rwlAERGgYkhQZKZ#_b
z<+bYvn^E#89gYqj4CSMPw7na(hEg2tx1}zujh?AWzUc0t)1)b}c}C(v7_7oQtHfmq
z<<s?5P&r*6wh!xzS<Ev-ug;83T`~m|{JD{to78RzQ`anAvr{Xx(dwjj<_POu8sSno
zW2?rr?&Wo=tx|shKJqg6cl?r(*|QQi#{RgN`RD^WsP7_LvLZ&D)r*-appk3NO3*Bg
z%UKDUZdRCTma;N5&+6sW!Q90O$1dSJs9}VzkuH4)^|pn5b9P#AguZCY$US>_l6N3F
zMrM3o8?%hXfcgY(xn%sAKPENv6+3uIiZB0ztHG3=jC$P^IOJ<r73=C9%|#KTUis?;
zjn=R~y4j%V>06F*@_~#(q2ZA<Q<T%S_1>@*_L7?@%oN^i;Cq-s8@N1X0vLVd$YsF>
zT=1ER%4~(HBqJs5ytbyn<sy}9vxYMp<r3Nm^;6?sl8suu;VM2%l%-+pimn`leZ*O(
zH+-nfUNlf<caUnRH7d$Ik~!WB6m89<dlD&+G)YN1G^Yy@_<1-q2S@$D#w1Ol*dxV4
z#U6qbW>M3mNmOw<mMUP1vrY<dRhsouZO-wkI`82RXoc@X7_7v7USSpHu@bA|e&?K|
zc1dGf#0qrLK9c#G9aiDij0=_mBE$z1whVv}8pk$JJ8L)3wL6(5*dtT^E90+grC`y_
zTAiQA&CKbz=`82QouAjE`Q~R3rnUvKcdl=$HSO~eF%DwJ%L<v_cJ{CyJr<FCwxYRt
z^1e6CLvtW7go)s=PX1Z<>$tvO*Az_jG?K$ny_?i!^!TVNBT-+&#z(RXu-DtYVVKs}
z>({1AB6@N<emF|I!$H_vtM7bm{MEG^v(FIJ=0EXv6+%%x9Ef_3o_Z*wI)z-4n|>BI
zJBXQgvG0;7p-~LVudOM_VNTzc1sbKn?ah{OZZv%`SU6SAKO5+gy2~e>X=^^IZZ7ID
z%G>9`?Q7D})c+B_ovc6@#Z|s)dHg%L-@{$CNXl=ar$vW3T?Au-wTALl6zCuTkd2s6
zaIE2Nf?fgZHXsoLwlC<YvtoNf@V2qV99{`*p`*Ku=#S)R*b8Z^8V-2p$#vy)`xJUX
zb^FuZt&Pp*C%YTJ-n+eVcS|`kBE6uvy|cfy+q`}M*7j}XrNbnZKuW{We#Cj+eA*Q_
zH@&2J7<D^`Y2bDHohGm%KsfzQU{kgIHJ{LcX15=9q83(`ABR@OTHzj-YiK#8Q)0%`
z<B&F&yb#lh86AZkhINDUUohhP9pd;WVecqH!N-b!_+gXa(y$ay+sfAas@yx^@FN+i
z!ZwY#O9Pm^g1P6-tTgjE+PQ+n5Y4*=mp?%He>PHoj7)s^48WP1D6f$P*Gb%fP=#C6
z`Bv+4^{I8oP?}lyW0YIuHoEA@Ai#fuC;0;$N{mJM(t$O$$IjRtd&GKR#X<_#!mR~l
z1io-_E*Hn8<G-dBvITd-CT!FA{P*KBc+AbpmkgX{pJgRs{s^c2z{`rdMCZ8f&3I-`
z%y9+Wg|Xv>6kIuBkl<m+1xOFyE5v?Qp}B?ir8Tair5KmAEN4|=V{@GsY`mEHaV1-f
z7bakEW;|YhV5C*b5GF{G(6=}-kn`2zrFfZ2SdTfsp5qg}4rE4p9r3sZOYN+NjAv%4
zL$&{}rGH|Ntp{(gP!M3fot(jzu`~Mk2T+|nj+C7TmaR!eA&d|!fQ&ocZ4|v}F#-4K
z1cF54f<6#6m{ei(EIk|!n%$wQ?BB^#(+9E(B2cyr2c1aMr4Bd^Xi2zH4rM@Ls=P31
zYJ_P}DtGHR?2nMUdw241NMIvPrWW^bMua`k#r|+l^MkTSpovu3t==$+l$ktBl#`@l
zc!aig?<hHJEJ;d^<Tprsm&Etz+x~I0-9@5-G$imesS=6!XY>jxrwjtz(FuxqG51h`
z^rlI?HcR4{XV{V2x0FdKcMv@#JyoDWu2rPs)Q`%IrYmIeJQFN-qO_S#lX`99Z!u+Z
z3&QYfl*S`LsUb!2xyPztF^AE;Y_0Md^U+67iP!k+eajf>b9o=LoQ5m?4V<Phe?ui_
zJe~Gi>Ffsp2@Dc^uzl=g4kh@<aq6ZXCH@-T!PrX+fVT&D<BUKc7E)h_o{!LHX%V`m
zxJ;pr&|_bkV;?P*ctOi*<`bt16QNhk+;}l7#<gkaeF!+a@sbX><@iFpLM1`@wvFE_
z`4d<y>7bMA5Fpl{=C7O7q!D^i(h(>xl2|1{e?&R^{ftT<lh`0Zd_RXyNiLUkCgxPR
zNu@0kcSsP&$!!w9An{8QcOe?39B!vblb=%8H%XAQ<qnBok<eN9Ju2;zpdXF5qhW+h
z=^q%C5TKE;Vy`X88WL5F0ltbdAwYhHet1NCVFgDlGMiyv|Ak6A^>e?FH<XRU3%`zh
zFp}S<IwD|@3?-7Vr@5zSc|qy^ZgX>Ef1`PCWB=#6c5ioU^Q^YNwYT5g+YYvL-k%c#
z<rwomCH!gQ&MlPwb|t0L%3JRb#ZfQ1Np>fcZdb7!575=OZ83{mj#u<n{D!ySIk-2x
GrT+pAi{Z%t

literal 0
HcmV?d00001

diff --git a/image_pro/tfRecord/__pycache__/__init__.cpython-36.pyc b/image_pro/tfRecord/__pycache__/__init__.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..acd70d6212ae83d5b3fae5227fce220a4d6fda52
GIT binary patch
literal 250
zcmXr!<>l%-@+``ffq~&M5W@izkmUfx#WFx5g&~D8has0Sijfh-W&*OAqL_ee=3oX*
zmY0k`C7O)4I6~5jQj_zGQh;p!6d)@lHNFHY<fqAei=hZ)Xc04zSjkYt3SxtZUkS-p
zF`>n&Ma3~bsd@Qj`7ti3#n~nK1u-S5#U;KzG2!_|*+wxgi6x2g0Y#|=Mfu68#l@L<
z>G8fkx+%s6x&|h?Mh1o!V0AGiX+dB&#KgyE=4F<|$LkeT-r}&y%}*)KNwovHv>4<H
G9!3DogF@#3

literal 0
HcmV?d00001

diff --git a/image_pro/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc b/image_pro/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..570936efb2c0e03693e276e21fb0ad73e18a506b
GIT binary patch
literal 2994
zcmbtWOK%)S5T5RN?)aI+iEI+WB1m8%v4Rt$2!ewY;{y*X5KyFzqse%?cXu|kGeb{1
zSzEIwM{*AM2O#kW_)C34;^Z4=E>P7o`v{6qB=@RodaHY?s=KScZ*I=b`R9H;X)Z4s
z#-GMiGhzHC^n4QpGZ=}DCecsYH1*SJTKZ}CoTkfWna!LRr0KCa=CK-#YHXhQYz9U?
zBM*$m!U<S1woeGi6L4p2wJg*?v!SCy?D-OiBjbo1(YDD5qc6OnRajua+-|{?Zn%&_
zKL<VE0HGHeLqi%p(xWfQAsLc_^vuGjM%GJOjqH~ropWdmX+cZV7#21MI>z$=dOiSw
zCOar8Bgb_74`Mr}jQ&Q41awwm6y%T&ao#Q{Ggt5p2Q(%;UFaTYEv>B`Qar3<I<%O}
zycfpMF6_dD@JwbFW<(0BjX~A68${Vm?q|F~l^ylN4p(*@?($e*gaSi7!DG<tcqY58
z(JoMnhopD$77QQ)Qs`ZPcWcRD=CXlD1b*y;TldoB85gqtL_2v>Z;P~Fe{|zoeQkaH
zasB#6{n7OstLxY5ARgD-aT>~ewF6e6{dM=`A9z&%{<zy&)GfQ)Fcp<;WwpXwDp$XB
zbFD#}uJ#cO^d5dU#2mc+_wCkaJ3r-I<XV2RwzJ9er!viU9;V{yj%@$HTd80>3?7UJ
zvMu!g)ofq+L6VAo7)J*j_nGUQq8G@)_j*;pLKy~8lI_W$ovhx8(^eSg8>_EVWueG<
zP??^r9NTaB!Uq>>VJjMN_AO7gcqsP-S57WPlyu;Z68Y&ZwNx6)K^60W55j&HbLCgT
zi}1O!_h991<)#S_x~UX1Xcpgw6LM5dwyP+V*DhV;$|>VgwetADioSob0I%vZEX6!p
znhU14g0+^TFz7s4qO<fYxkNqNqX%D{>WCU{c^noVv0C_XtGuL5_?L63%zl_*)EVqR
z*}I`^b=Adw_>{xZM>4`23wPmsBFRN4Q_(QRB@Et0b7;Vu%qU;KALMY$s@9HTo`ij#
zD?i!m2Rut#-F#~MwejWR;pBKE2Jx0=iBEm#o_TQf%{H!WHf(Vf42Ta<Tmhlnyt~(K
z$6S1fD$HMq9cQK%>U%i%8aKGSO47x;_z3=;)S*&aSh3h@&@qSaLC=2yftA~%1wh;}
z(fDHaEC4QZkQ8*&c-#hbGy!w0LqZJcz}W41jIJ2HTF*xu0$j1jdO)Bl{b;of+L?0o
z(5);ula)j7=gSWPF#s!E#3U;VWmh>}9(B4>Amk|fdBkLQ0!xkAGB;}3zU-z+5T)wu
zIIG83u52m7C{gZLRgEm%#}eEE^m$NLlnFH<R?vf6Rf6($kp*!Cuwp+8<2@*?-F?Y(
zt!mViKL|z{(u8p<%)(YA_nQl41x>uWRcSU|s9&r&aUDE~RS-=->UV;&tjjvXdz;QN
z!9%9~tieRS48kDum|yT)q#lKC&Qpgjo8ohrYatqA!@I>Xc3qW6Mw?VT2k2jt5zOg^
zi^n{Nj{d`QBPN^Txsua*Hm1nP+=vxk;kK`}In1t#-RLomfHpw`=EiJh0dEn_TQ;&8
zGnoa91&pj2%VwmoBh4b)`t}<{V9;YCLf5#}|0g1&fCZ-}B7wE@IwI1!s6kfD!V-v4
z^XIEqs*&^|Oiw7}{#5<8LivoPol`Q1Q-*?QEY6_7FBpNJJmNfvh9fYObjAy$GjS0G
z!q|v6lQIc)><BU%G`C@p<JJtS0p0|7w+`-{CVU75$xghD;xdYNK&T5E!I8IeAg)O*
zc^xi;gb!rH(h-UG@Rj#bVEMw4YY?4;joA4e^fGMXod<ZfJp!zB<&DBFao3uHT#djJ
zNUl6;AO`G#@-Dvxs*C&vePgN`boO0=-;e;}SZe~)bqfOi7<F`_Ys@HL@+40Me9SyA
z@y!MTf{zfIWQwM_I@1qj7N;_fcH!FuN6M+5W)@2g_Zkkg!>OAbpNsPD^i#GM^9_te
Ra~B*6{Cv;9;;;DY{{o=%+XDaq

literal 0
HcmV?d00001

diff --git a/image_pro/tfRecord/decode.py b/image_pro/tfRecord/decode.py
new file mode 100644
index 0000000..d2a5546
--- /dev/null
+++ b/image_pro/tfRecord/decode.py
@@ -0,0 +1,129 @@
+#NOTE: refer to this url: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/how_tos/reading_data/fully_connected_reader.py
+import tensorflow as tf
+import numpy as np
+import os
+import matplotlib.pyplot as plt
+
+
+
+
+
+
+
+def tfrec_data_input_fn(filenames, num_epochs=1, batch_size=16, shuffle=False):
+    
+    def _input_fn():
+        def _parse_record(tf_record):
+            features = {
+                'image': tf.FixedLenFeature([], dtype=tf.string),
+                'label': tf.FixedLenFeature([], dtype=tf.int64)
+            }
+            record = tf.parse_single_example(tf_record, features)
+
+            image_raw = tf.decode_raw(record['image'], tf.uint8)
+            label = tf.one_hot(tf.cast(record['label'], tf.int32), depth=2)
+            
+            def _normalize(image_x):
+                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+                image = tf.cast(image_x, tf.float32) * (1. / 255)-0.5
+                return image
+            
+            image_raw = _normalize(image_raw)
+            '''image_raw = tf.reshape(image_raw, (32,32,3))'''#For vizualizing 
+            return { 'image': image_raw}, label
+        
+        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
+        dataset = tf.data.TFRecordDataset(filenames)
+        dataset = dataset.map(_parse_record)
+        
+        dataset = dataset.repeat(num_epochs)
+        dataset = dataset.batch(batch_size)
+
+        iterator = dataset.make_initializable_iterator()
+        '''features, labels = iterator.get_next()'''
+
+        return iterator
+    
+    return _input_fn
+
+
+def read_file(x_filename, x_capacity=800):
+    record_iterator = tf.python_io.tf_record_iterator(path=x_filename)
+
+    example = tf.train.Example()
+    for str_rec in record_iterator:
+        example.ParseFromString(str_rec)
+        height = int(example.features.feature['height']
+                                 .int64_list
+                                 .value[0])
+
+        width = int(example.features.feature['width']
+                                .int64_list
+                                .value[0])
+    
+
+        img_string = (example.features.feature['image']
+                                  .bytes_list
+                                  .value[0])
+
+        label = (example.features.feature['label'].int64_list.value[0])
+
+
+    return 0
+
+
+def tfrec_data_catvdog(filenames, num_epochs=1, batch_size=16):
+    
+    def _input_fn():
+        def _parse_record(tf_record):
+            features = {
+                'image': tf.FixedLenFeature([], dtype=tf.string),
+            }
+            record = tf.parse_single_example(tf_record, features)
+
+            image_raw = tf.decode_raw(record['image'], tf.uint8)
+            
+            
+            def _normalize(image):
+                """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+                image = tf.cast(image, tf.float32) * (1. / 255)-0.5
+                return image
+
+            image_n = _normalize(image_raw)
+            
+            
+
+            return { 'image': image_n }
+        
+        # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
+        dataset = tf.data.TFRecordDataset(filenames)
+        dataset = dataset.map(_parse_record)
+
+        '''dataset = dataset.repeat(num_epochs)'''
+        '''dataset = dataset.batch(batch_size)'''
+
+        iterator = dataset.make_one_shot_iterator()
+        features= iterator.get_next()
+
+        return features
+    
+    return _input_fn
+
+
+
+
+
+'''read_file("./tftest.tfrecords")
+
+tfrec_dev_input_fn = tfrec_data_input_fn(["tftest.tfrecords"])
+features, labels = tfrec_dev_input_fn()
+
+with tf.Session() as sess:
+    for step in range(2):
+        img, label = sess.run([features['image'], labels])
+        img = np.reshape(img, (16,64,64,3))
+    
+        for i in range(16):
+            plt.imshow(img[i])
+            plt.show()'''
+        
\ No newline at end of file
diff --git a/image_pro/utils_data/__init__.py b/image_pro/utils_data/__init__.py
new file mode 100644
index 0000000..ab28e3f
--- /dev/null
+++ b/image_pro/utils_data/__init__.py
@@ -0,0 +1 @@
+#NOTE
\ No newline at end of file
diff --git a/image_pro/utils_data/colors/__init__.py b/image_pro/utils_data/colors/__init__.py
new file mode 100644
index 0000000..ab28e3f
--- /dev/null
+++ b/image_pro/utils_data/colors/__init__.py
@@ -0,0 +1 @@
+#NOTE
\ No newline at end of file
diff --git a/image_pro/utils_data/colors/colors.py b/image_pro/utils_data/colors/colors.py
new file mode 100644
index 0000000..a209e53
--- /dev/null
+++ b/image_pro/utils_data/colors/colors.py
@@ -0,0 +1,117 @@
+import cv2
+
+#############
+#Color change
+#############
+#Changing color space:
+def change_color_gray(frame):
+    change = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+    return change
+
+def change_color_hsv(frame):
+    change = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
+    return change
+
+def change_color_rgb(frame):
+    change = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
+    return change
+
+#Object tracking: using a specified range of color
+#Ex: lower_blue = np.array([110,50,50]); upper_blue = np.array([130,255,255])
+#lower_col & upper_col have to be numpy arrays
+
+def track_color(frame, lower_col, upper_col):
+    hsv = change_color_hsv(frame)
+    mask = cv2.inRange(hsv, lower_col, upper_col)
+    img = cv2.bitwise_and(frame,frame,mask=mask)
+    return img
+
+#############
+#Thresholding
+#############
+
+#Images processed in this part must be grayscale
+#Ex: if x>value_min --> white else black
+# value_aft is given if pixel > value_thr
+
+def thresh_bin(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_BINARY)
+    return img
+
+def thresh_bin_inv(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_BINARY_INV)
+    return img
+
+def thresh_truc(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TRUNC)
+    return img
+
+def thresh_tozero(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TOZERO)
+    return img
+
+def thresh_tozero_inv(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_TOZERO_INV)
+    return img
+
+#Adaptive thresholding
+#block_h: means block size height, block_w means block size width
+
+##With binary thresholding
+
+def thresh_adapgauss_bin(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_BINARY, block_h, block_w )
+
+    return img
+
+def thresh_adapgauss_bininv(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_BINARY_INV, block_h, block_w )
+
+    return img
+
+def thresh_adapgauss_tunc(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_TRUNC, block_h, block_w )
+
+    return img
+
+def thresh_adapgauss_toz(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_TOZERO, block_h, block_w )
+
+    return img
+
+def thresh_adapgauss_tozinv(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+                                cv2.THRESH_TOZERO_INV, block_h, block_w )
+
+    return img
+
+def thresh_adapmean_bin(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
+                                cv2.THRESH_BINARY, block_h,block_w)
+    return img
+
+def thresh_adapmean_trunc(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
+                                cv2.THRESH_TRUNC, block_w, block_h)
+    return img
+
+def thresh_adapmean_toz(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
+                                cv2.THRESH_TOZERO, block_h, block_w)
+    return img
+
+def thresh_adapmean_tozinv(frame, value_aft, block_h, block_w):
+    img = cv2.adaptiveThreshold(frame, value_aft, cv2.ADAPTIVE_THRESH_MEAN_C,\
+                                cv2.THRESH_TOZERO_INV, block_h, block_w)
+    return img
+
+
+#################
+#Otsu thresholding
+##################
+
+#TODO: ret2,th2 = cv2.threshold(img,0,255,cv2.THRESH_BINARY+cv2.THRESH_OTSU)
\ No newline at end of file
diff --git a/image_pro/utils_data/contours/__init__.py b/image_pro/utils_data/contours/__init__.py
new file mode 100644
index 0000000..ab28e3f
--- /dev/null
+++ b/image_pro/utils_data/contours/__init__.py
@@ -0,0 +1 @@
+#NOTE
\ No newline at end of file
diff --git a/image_pro/utils_data/contours/contours.py b/image_pro/utils_data/contours/contours.py
new file mode 100644
index 0000000..0b66f29
--- /dev/null
+++ b/image_pro/utils_data/contours/contours.py
@@ -0,0 +1,128 @@
+import numpy as np
+import cv2
+'''from ..colors.colors import thresh_bin'''
+import matplotlib.pyplot as plt
+
+####################
+# Canny contours
+####################
+def canny_cont(frame, threch1, threch2):
+    img_save = frame
+    edges = cv2.Canny(frame, threch1,threch2)
+    return edges
+
+
+###################
+# Thresholding
+####################
+
+def thresh_bin(frame, value_thr, value_aft):
+    _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_BINARY)
+    return img
+
+def find_contours(frame, approx=False):
+    tresh = thresh_bin(frame, 70, 255)
+    if approx:
+        image, contours, hierarchy = cv2.findContours(tresh, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
+    else:
+        image, contours, hierarchy = cv2.findContours(tresh, cv2.RETR_TREE, cv2.CHAIN_APPROX_NONE)
+
+    return image, contours, hierarchy
+    
+    #image: source image
+    # contours: contour retrieval mode
+    # hierarchy: contour approximation method
+
+def draw_all_contours(frame):
+    save_img = frame
+    tresh, contours, _ = find_contours(frame)
+    img = cv2.drawContours(tresh, contours, -1,(0,255))
+    return img, save_img
+
+def draw_contour(frame, pos):
+    #pos is an int begining from 0
+    save_img = frame
+    tresh, contours, _ = find_contours(frame)
+    image = cv2.drawContours(tresh, contours, pos, (0,255))
+    return image, save_img
+
+def draw_contour_approx(frame):
+    save_img = frame
+    tresh, contours, _ = find_contours(frame, approx=True)
+    image = cv2.drawContours(tresh, contours, -1, (0,255))
+    return image, save_img
+
+####################
+# Contours Features
+####################
+
+def get_moment(frame, index):
+    _, contours, _ = find_contours(frame, approx=True)
+    cnt = contours[index]
+    M = cv2.moments(cnt)
+    return M          #M is a dict of all the moments
+
+
+def get_contour_area(contour):
+    area = cv2.contourArea(contour)
+    return area
+
+def get_perimeter(contour):
+    perimeter = cv2.arcLength(contour)
+    return perimeter
+
+def get_approx_cont(contour):
+    epsilon = 0.1*cv2.arcLength(contour,True)
+    approx = cv2.approxPolyDP(contour,epsilon,True)
+    return approx
+
+
+####################
+# Bounding Rectangle
+####################
+def bounding_rec(frame):
+    _, contours,_ = find_contours(frame, approx=True)
+    image = frame
+    for c in contours:
+        x,y,w,h = cv2.boundingRect(c)
+        approx = get_approx_cont(c)
+        if len(approx) >= 4:
+            image = cv2.rectangle(image,(x,y),(x+w,y+h),(0,255,0))
+    
+    return image
+
+def draw_rec(frame, index):
+    _,contours,_ = find_contours(frame, approx=True)
+    cnt = contours
+    rect = cv2.minAreaRect(cnt[0])
+    box = cv2.boxPoints(rect)
+    box = np.int0(box)
+    image = cv2.drawContours(frame, [box],0,(0,0,255),2)
+    return image
+
+
+########################
+# Contours Properties
+########################
+def get_ratio(contour):
+    x,y,w,h = cv2.boundingRect(contour)
+    ratio = float(w)/h
+    return ratio
+
+def get_extent(contour):
+    area = get_contour_area(contour)
+    x,y,w,h = cv2.boundingRect(cnt)
+    rect_area = w*h
+    extent = float(area)/rect_area
+    return extent
+
+def mask_pix(frame_gray):
+    _,contours,_ = find_contours(frame_gray)
+    mask = np.zeros(frame_gray.shape,np.uint8)
+    cv2.drawContours(mask,contours,0,255)
+    pixelpoints = cv2.findNonZero(mask)
+    return pixelpoints
+
+
+#Aspect ratio: It is the ratio of width to height of bounding rect of the object
+
diff --git a/image_pro/utils_data/geotransf/__init__.py b/image_pro/utils_data/geotransf/__init__.py
new file mode 100644
index 0000000..ab28e3f
--- /dev/null
+++ b/image_pro/utils_data/geotransf/__init__.py
@@ -0,0 +1 @@
+#NOTE
\ No newline at end of file
diff --git a/image_pro/utils_data/geotransf/geotransf.py b/image_pro/utils_data/geotransf/geotransf.py
new file mode 100644
index 0000000..019618d
--- /dev/null
+++ b/image_pro/utils_data/geotransf/geotransf.py
@@ -0,0 +1,81 @@
+import cv2
+import numpy as np
+import matplotlib.pyplot as plt
+#Let's consider M the translation matrix. Given a translation vector (tx,ty),
+# the matrix is defined as follows: M = [1 0 tx]
+#                                       [0 1 ty]
+
+
+##################
+# Translation
+##################
+def translate(frame, tx, ty):
+    dim = frame.shape
+    M = np.float32([[1,0,tx],[0,1,ty]])
+    img = cv2.warpAffine(frame, M, dim)
+    return img
+
+################
+# Rotation
+################
+#See reference: http://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_imgproc/py_geometric_transformations/py_geometric_transformations.html
+#Angle is in degree
+def rotate(frame, angle):
+    dim = (int(frame.shape[0]/2), int(frame.shape[1]/2))
+    M = cv2.getRotationMatrix2D(dim, angle, 1) #Last argument is for scale
+    img = cv2.warpAffine(frame, M, dim)
+    return img
+
+#######################
+# Affine transformation
+#######################
+# For this func , we need three points from the original image
+# And the corresponding points in the output image 
+#list1: list of tuples of input image: 3 tuples// Same for list2 (output image) 
+def affine(frame, list1, list2):
+    '''assert type(list1) == list and len(list1)==3'''
+    col,row,_ = frame.shape
+    sub1, sub2 = [], []
+    for i in list1:
+        sub1.append([i[0],i[1]])
+    for j in list2:
+        sub2.append([j[0], j[1]])
+    ps1 = np.float32(sub1)
+    ps2 = np.float32(sub2)
+    M = cv2.getAffineTransform(ps1,ps2)
+    img = cv2.warpAffine(frame, M, dsize=(col,row))
+    return img
+
+
+
+#######################
+# Perspective transformation
+#######################
+
+def perspec(frame, list1, list2):
+    '''assert type(list1) == list and len(list1)==4'''
+    col,row,_ = frame.shape
+    sub1, sub2 = [], []
+    for i in list1:
+        sub1.append([i[0],i[1]])
+    for j in list2:
+        sub2.append([j[0], j[1]])
+    ps1 = np.float32(sub1)
+    ps2 = np.float32(sub2)
+    M = cv2.getPerspectiveTransform(ps1,ps2)
+    img = cv2.warpPerspective(frame, M, dsize=(col,row))
+
+    return img
+
+'''a = cv2.imread("C:/Users/User12/Desktop/Work/train/train/cat.0.jpg", cv2.IMREAD_UNCHANGED)
+b = perspec(a, [(50,50),(300,50),(50,300), (300,300)], [(0,0),(250,0),(0,250),(250,250)])
+plt.imshow(b)
+plt.show()'''
+
+
+#######################
+# Morphological transformation
+#######################
+#TODO:
+
+
diff --git a/image_pro/utils_data/norma/__init__.py b/image_pro/utils_data/norma/__init__.py
new file mode 100644
index 0000000..ab28e3f
--- /dev/null
+++ b/image_pro/utils_data/norma/__init__.py
@@ -0,0 +1 @@
+#NOTE
\ No newline at end of file
diff --git a/image_pro/utils_data/norma/norma.py b/image_pro/utils_data/norma/norma.py
new file mode 100644
index 0000000..e6cd714
--- /dev/null
+++ b/image_pro/utils_data/norma/norma.py
@@ -0,0 +1,22 @@
+import tensorflow as tf
+import cv2
+
+def normalize(image, label):
+  """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
+  image = tf.cast(image, tf.float32) * (1. / 255) - 0.5
+  return image, label
+
+####################
+# Image Gradient
+####################
+
+def laplacian(frame):
+    img = cv2.Laplacian(frame, cv2.CV_64F)
+    return img
+
+def sobelX(frame, kernel_size):
+    img = cv2.Sobel(frame, cv2.CV_64F, 1, 0, ksize=kernel_size)
+    return img
+
+def sobelY(frame, kernel_size):
+    img = cv2.Sobel(frame, cv2.CV_64F, 0,1,ksize=kernel_size)
\ No newline at end of file
diff --git a/image_pro/utils_data/zooming/__init__.py b/image_pro/utils_data/zooming/__init__.py
new file mode 100644
index 0000000..ab28e3f
--- /dev/null
+++ b/image_pro/utils_data/zooming/__init__.py
@@ -0,0 +1 @@
+#NOTE
\ No newline at end of file
diff --git a/image_pro/utils_data/zooming/zoom.py b/image_pro/utils_data/zooming/zoom.py
new file mode 100644
index 0000000..ab28e3f
--- /dev/null
+++ b/image_pro/utils_data/zooming/zoom.py
@@ -0,0 +1 @@
+#NOTE
\ No newline at end of file

From 8a51cd8107e5d0a8617c640dcce0319c5b01e535 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:52:18 +0200
Subject: [PATCH 25/28] Rename utils_data/histogram/histogram.py to
 image_pro/utils_data/histogram/histogram.py

---
 {utils_data => image_pro/utils_data}/histogram/histogram.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
 rename {utils_data => image_pro/utils_data}/histogram/histogram.py (97%)

diff --git a/utils_data/histogram/histogram.py b/image_pro/utils_data/histogram/histogram.py
similarity index 97%
rename from utils_data/histogram/histogram.py
rename to image_pro/utils_data/histogram/histogram.py
index 89b4fa4..7ef0af8 100644
--- a/utils_data/histogram/histogram.py
+++ b/image_pro/utils_data/histogram/histogram.py
@@ -13,4 +13,4 @@ def plot_bgr_histogram(frame, mask=None, histosize=[256], ranges=[0, 256]):
         histr = cv2.calcHist([frame], [i], mask, [256], ranges)
         plt.plot(histr, color=col)
         plt.xlim([0,256])
-    plt.show()
\ No newline at end of file
+    plt.show()

From 0b89416bd54744bd5c3b3f596e9eb594dad7c357 Mon Sep 17 00:00:00 2001
From: Alaa El Bouchti <alaa_23@hotmail.fr>
Date: Thu, 3 May 2018 15:53:05 +0200
Subject: [PATCH 26/28] Rename utils_data/histogram/__init__.py to
 image_pro/utils_data/histogram/__init__.py

---
 image_pro/utils_data/histogram/__init__.py | 1 +
 utils_data/histogram/__init__.py           | 0
 2 files changed, 1 insertion(+)
 create mode 100644 image_pro/utils_data/histogram/__init__.py
 delete mode 100644 utils_data/histogram/__init__.py

diff --git a/image_pro/utils_data/histogram/__init__.py b/image_pro/utils_data/histogram/__init__.py
new file mode 100644
index 0000000..d3f5a12
--- /dev/null
+++ b/image_pro/utils_data/histogram/__init__.py
@@ -0,0 +1 @@
+
diff --git a/utils_data/histogram/__init__.py b/utils_data/histogram/__init__.py
deleted file mode 100644
index e69de29..0000000

From 4967fe3b922c9b5504d3d89179b7e09d55d0baac Mon Sep 17 00:00:00 2001
From: a-hilaly <hilalyamine@gmail.com>
Date: Thu, 3 May 2018 21:33:33 +0200
Subject: [PATCH 27/28] Add makefile & requirements.txt

---
 Makefile                                      |  56 ++++++++++++++++++
 image_pro/tfRecord/Tfrecord.py                |   9 +--
 .../__pycache__/Tfrecord.cpython-36.pyc       | Bin 4269 -> 0 bytes
 .../__pycache__/__init__.cpython-36.pyc       | Bin 250 -> 0 bytes
 .../decode_tfrecord.cpython-36.pyc            | Bin 2994 -> 0 bytes
 requirements.txt                              |   4 ++
 CNN-Xray.py => scripts/CNN-Xray.py            |   0
 NN2-Xray.py => scripts/NN2-Xray.py            |   0
 NN3-Xray..py => scripts/NN3-Xray..py          |   0
 Tfrecord2.py => scripts/Tfrecord2.py          |   0
 .../decode_tfrecord.py                        |   0
 image_pro/setup.py => setup.py                |  19 +++---
 12 files changed, 70 insertions(+), 18 deletions(-)
 create mode 100644 Makefile
 delete mode 100644 image_pro/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
 delete mode 100644 image_pro/tfRecord/__pycache__/__init__.cpython-36.pyc
 delete mode 100644 image_pro/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
 create mode 100644 requirements.txt
 rename CNN-Xray.py => scripts/CNN-Xray.py (100%)
 rename NN2-Xray.py => scripts/NN2-Xray.py (100%)
 rename NN3-Xray..py => scripts/NN3-Xray..py (100%)
 rename Tfrecord2.py => scripts/Tfrecord2.py (100%)
 rename decode_tfrecord.py => scripts/decode_tfrecord.py (100%)
 rename image_pro/setup.py => setup.py (69%)

diff --git a/Makefile b/Makefile
new file mode 100644
index 0000000..f539ddc
--- /dev/null
+++ b/Makefile
@@ -0,0 +1,56 @@
+# Makefile
+
+# Shell
+SHELL := /bin/bash
+#Project
+PROJECT = "image_pro"
+#Version
+RELEASE = "0.0.1"
+# Path to source directory
+PROJECTPATH := .
+# OS machine
+OS_TYPE = 'uname -a'
+# CPP EXTENSIONS
+CPP_EXT = ".cpp"
+
+# SETUP
+SETUPFILE = "setup.py"
+
+# VENV
+
+_virtualenv:
+	# create virtual env
+	virtualenv _virtualenv
+	_virtualenv/bin/pip install --upgrade pip
+	_virtualenv/bin/pip install --upgrade setuptools
+
+_use_env:
+	# use virtual env
+	source _virtualenv/bin/activate
+
+##### PYTHON
+
+prebuild:
+	@echo "Preparing build"
+	@echo "Installing requirements"
+	pip install -r requirements.txt
+
+setup:
+	echo "Setup..."
+	sudo python setup.py install
+
+test:
+	echo "Test"
+	python setup.py test
+
+clean:
+	rm -f MANIFEST
+	rm -rf build dist
+
+deactivate_env:
+	deactivate
+
+
+all: prebuild setup
+
+.PHONY: clean
\ No newline at end of file
diff --git a/image_pro/tfRecord/Tfrecord.py b/image_pro/tfRecord/Tfrecord.py
index 718719b..1451b4e 100644
--- a/image_pro/tfRecord/Tfrecord.py
+++ b/image_pro/tfRecord/Tfrecord.py
@@ -1,7 +1,7 @@
 from random import shuffle, seed
 import os
 import glob
-from .utils_data.contours.contours import draw_all_contours
+from image_pro.utils_data.contours.contours import draw_all_contours
 import numpy as np
 
 import cv2
@@ -11,13 +11,6 @@
 import sys
 
 
-
-
-
-
-
-
-
 # This function returns the different paths and corresponding labels
 def shuffling_data(data_path, labels):
     c = list(zip(data_path, labels))
diff --git a/image_pro/tfRecord/__pycache__/Tfrecord.cpython-36.pyc b/image_pro/tfRecord/__pycache__/Tfrecord.cpython-36.pyc
deleted file mode 100644
index 6446f71bb9044e55ae8e57687425f3d41c491da8..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 4269
zcmb_fTW=gm74E9OPS3^TvEw+q3mUB!!31z_%WY-CI<d1DvL>>fjc9opb*6g8)0tbU
z+lgcK@G=4M5(H?))BXVX6^MVpbG_~h`3rf0?^Ms&&TNziM7R2M-A?tXQ{VaO^o?@a
z55IUB{N|Ejd}W-wOq4&yll%oj7=rbTJ_}f%2b@uz_sqc5Z!56$+YTK4b^}+xy`Z4q
ze&7o;C<-em2|FkYN4U?8pdvg`z<WWg2wxPRv!E(UqKuYBQ4tGx*F;q;;=LqlVhQhM
zu`Dj&eL?U&qjB*i>@pf$xyj*CyWNYFl|+%)Iaye{lFG0<xN`Z*<+XU!IaxwOl%&_D
zQa5)0jrFxSO@YhCW8t}tC%Fuf8Y#<+j7|82jd{v5ycxCeKN<&oY>utW%&a!q!FC#^
zvU=SlRpv=|Bx%lutqLMc!{#VV50%pkA4a`Iv6ix^HE?x9B>rwlAERGgYkhQZKZ#_b
z<+bYvn^E#89gYqj4CSMPw7na(hEg2tx1}zujh?AWzUc0t)1)b}c}C(v7_7oQtHfmq
z<<s?5P&r*6wh!xzS<Ev-ug;83T`~m|{JD{to78RzQ`anAvr{Xx(dwjj<_POu8sSno
zW2?rr?&Wo=tx|shKJqg6cl?r(*|QQi#{RgN`RD^WsP7_LvLZ&D)r*-appk3NO3*Bg
z%UKDUZdRCTma;N5&+6sW!Q90O$1dSJs9}VzkuH4)^|pn5b9P#AguZCY$US>_l6N3F
zMrM3o8?%hXfcgY(xn%sAKPENv6+3uIiZB0ztHG3=jC$P^IOJ<r73=C9%|#KTUis?;
zjn=R~y4j%V>06F*@_~#(q2ZA<Q<T%S_1>@*_L7?@%oN^i;Cq-s8@N1X0vLVd$YsF>
zT=1ER%4~(HBqJs5ytbyn<sy}9vxYMp<r3Nm^;6?sl8suu;VM2%l%-+pimn`leZ*O(
zH+-nfUNlf<caUnRH7d$Ik~!WB6m89<dlD&+G)YN1G^Yy@_<1-q2S@$D#w1Ol*dxV4
z#U6qbW>M3mNmOw<mMUP1vrY<dRhsouZO-wkI`82RXoc@X7_7v7USSpHu@bA|e&?K|
zc1dGf#0qrLK9c#G9aiDij0=_mBE$z1whVv}8pk$JJ8L)3wL6(5*dtT^E90+grC`y_
zTAiQA&CKbz=`82QouAjE`Q~R3rnUvKcdl=$HSO~eF%DwJ%L<v_cJ{CyJr<FCwxYRt
z^1e6CLvtW7go)s=PX1Z<>$tvO*Az_jG?K$ny_?i!^!TVNBT-+&#z(RXu-DtYVVKs}
z>({1AB6@N<emF|I!$H_vtM7bm{MEG^v(FIJ=0EXv6+%%x9Ef_3o_Z*wI)z-4n|>BI
zJBXQgvG0;7p-~LVudOM_VNTzc1sbKn?ah{OZZv%`SU6SAKO5+gy2~e>X=^^IZZ7ID
z%G>9`?Q7D})c+B_ovc6@#Z|s)dHg%L-@{$CNXl=ar$vW3T?Au-wTALl6zCuTkd2s6
zaIE2Nf?fgZHXsoLwlC<YvtoNf@V2qV99{`*p`*Ku=#S)R*b8Z^8V-2p$#vy)`xJUX
zb^FuZt&Pp*C%YTJ-n+eVcS|`kBE6uvy|cfy+q`}M*7j}XrNbnZKuW{We#Cj+eA*Q_
zH@&2J7<D^`Y2bDHohGm%KsfzQU{kgIHJ{LcX15=9q83(`ABR@OTHzj-YiK#8Q)0%`
z<B&F&yb#lh86AZkhINDUUohhP9pd;WVecqH!N-b!_+gXa(y$ay+sfAas@yx^@FN+i
z!ZwY#O9Pm^g1P6-tTgjE+PQ+n5Y4*=mp?%He>PHoj7)s^48WP1D6f$P*Gb%fP=#C6
z`Bv+4^{I8oP?}lyW0YIuHoEA@Ai#fuC;0;$N{mJM(t$O$$IjRtd&GKR#X<_#!mR~l
z1io-_E*Hn8<G-dBvITd-CT!FA{P*KBc+AbpmkgX{pJgRs{s^c2z{`rdMCZ8f&3I-`
z%y9+Wg|Xv>6kIuBkl<m+1xOFyE5v?Qp}B?ir8Tair5KmAEN4|=V{@GsY`mEHaV1-f
z7bakEW;|YhV5C*b5GF{G(6=}-kn`2zrFfZ2SdTfsp5qg}4rE4p9r3sZOYN+NjAv%4
zL$&{}rGH|Ntp{(gP!M3fot(jzu`~Mk2T+|nj+C7TmaR!eA&d|!fQ&ocZ4|v}F#-4K
z1cF54f<6#6m{ei(EIk|!n%$wQ?BB^#(+9E(B2cyr2c1aMr4Bd^Xi2zH4rM@Ls=P31
zYJ_P}DtGHR?2nMUdw241NMIvPrWW^bMua`k#r|+l^MkTSpovu3t==$+l$ktBl#`@l
zc!aig?<hHJEJ;d^<Tprsm&Etz+x~I0-9@5-G$imesS=6!XY>jxrwjtz(FuxqG51h`
z^rlI?HcR4{XV{V2x0FdKcMv@#JyoDWu2rPs)Q`%IrYmIeJQFN-qO_S#lX`99Z!u+Z
z3&QYfl*S`LsUb!2xyPztF^AE;Y_0Md^U+67iP!k+eajf>b9o=LoQ5m?4V<Phe?ui_
zJe~Gi>Ffsp2@Dc^uzl=g4kh@<aq6ZXCH@-T!PrX+fVT&D<BUKc7E)h_o{!LHX%V`m
zxJ;pr&|_bkV;?P*ctOi*<`bt16QNhk+;}l7#<gkaeF!+a@sbX><@iFpLM1`@wvFE_
z`4d<y>7bMA5Fpl{=C7O7q!D^i(h(>xl2|1{e?&R^{ftT<lh`0Zd_RXyNiLUkCgxPR
zNu@0kcSsP&$!!w9An{8QcOe?39B!vblb=%8H%XAQ<qnBok<eN9Ju2;zpdXF5qhW+h
z=^q%C5TKE;Vy`X88WL5F0ltbdAwYhHet1NCVFgDlGMiyv|Ak6A^>e?FH<XRU3%`zh
zFp}S<IwD|@3?-7Vr@5zSc|qy^ZgX>Ef1`PCWB=#6c5ioU^Q^YNwYT5g+YYvL-k%c#
z<rwomCH!gQ&MlPwb|t0L%3JRb#ZfQ1Np>fcZdb7!575=OZ83{mj#u<n{D!ySIk-2x
GrT+pAi{Z%t

diff --git a/image_pro/tfRecord/__pycache__/__init__.cpython-36.pyc b/image_pro/tfRecord/__pycache__/__init__.cpython-36.pyc
deleted file mode 100644
index acd70d6212ae83d5b3fae5227fce220a4d6fda52..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 250
zcmXr!<>l%-@+``ffq~&M5W@izkmUfx#WFx5g&~D8has0Sijfh-W&*OAqL_ee=3oX*
zmY0k`C7O)4I6~5jQj_zGQh;p!6d)@lHNFHY<fqAei=hZ)Xc04zSjkYt3SxtZUkS-p
zF`>n&Ma3~bsd@Qj`7ti3#n~nK1u-S5#U;KzG2!_|*+wxgi6x2g0Y#|=Mfu68#l@L<
z>G8fkx+%s6x&|h?Mh1o!V0AGiX+dB&#KgyE=4F<|$LkeT-r}&y%}*)KNwovHv>4<H
G9!3DogF@#3

diff --git a/image_pro/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc b/image_pro/tfRecord/__pycache__/decode_tfrecord.cpython-36.pyc
deleted file mode 100644
index 570936efb2c0e03693e276e21fb0ad73e18a506b..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 2994
zcmbtWOK%)S5T5RN?)aI+iEI+WB1m8%v4Rt$2!ewY;{y*X5KyFzqse%?cXu|kGeb{1
zSzEIwM{*AM2O#kW_)C34;^Z4=E>P7o`v{6qB=@RodaHY?s=KScZ*I=b`R9H;X)Z4s
z#-GMiGhzHC^n4QpGZ=}DCecsYH1*SJTKZ}CoTkfWna!LRr0KCa=CK-#YHXhQYz9U?
zBM*$m!U<S1woeGi6L4p2wJg*?v!SCy?D-OiBjbo1(YDD5qc6OnRajua+-|{?Zn%&_
zKL<VE0HGHeLqi%p(xWfQAsLc_^vuGjM%GJOjqH~ropWdmX+cZV7#21MI>z$=dOiSw
zCOar8Bgb_74`Mr}jQ&Q41awwm6y%T&ao#Q{Ggt5p2Q(%;UFaTYEv>B`Qar3<I<%O}
zycfpMF6_dD@JwbFW<(0BjX~A68${Vm?q|F~l^ylN4p(*@?($e*gaSi7!DG<tcqY58
z(JoMnhopD$77QQ)Qs`ZPcWcRD=CXlD1b*y;TldoB85gqtL_2v>Z;P~Fe{|zoeQkaH
zasB#6{n7OstLxY5ARgD-aT>~ewF6e6{dM=`A9z&%{<zy&)GfQ)Fcp<;WwpXwDp$XB
zbFD#}uJ#cO^d5dU#2mc+_wCkaJ3r-I<XV2RwzJ9er!viU9;V{yj%@$HTd80>3?7UJ
zvMu!g)ofq+L6VAo7)J*j_nGUQq8G@)_j*;pLKy~8lI_W$ovhx8(^eSg8>_EVWueG<
zP??^r9NTaB!Uq>>VJjMN_AO7gcqsP-S57WPlyu;Z68Y&ZwNx6)K^60W55j&HbLCgT
zi}1O!_h991<)#S_x~UX1Xcpgw6LM5dwyP+V*DhV;$|>VgwetADioSob0I%vZEX6!p
znhU14g0+^TFz7s4qO<fYxkNqNqX%D{>WCU{c^noVv0C_XtGuL5_?L63%zl_*)EVqR
z*}I`^b=Adw_>{xZM>4`23wPmsBFRN4Q_(QRB@Et0b7;Vu%qU;KALMY$s@9HTo`ij#
zD?i!m2Rut#-F#~MwejWR;pBKE2Jx0=iBEm#o_TQf%{H!WHf(Vf42Ta<Tmhlnyt~(K
z$6S1fD$HMq9cQK%>U%i%8aKGSO47x;_z3=;)S*&aSh3h@&@qSaLC=2yftA~%1wh;}
z(fDHaEC4QZkQ8*&c-#hbGy!w0LqZJcz}W41jIJ2HTF*xu0$j1jdO)Bl{b;of+L?0o
z(5);ula)j7=gSWPF#s!E#3U;VWmh>}9(B4>Amk|fdBkLQ0!xkAGB;}3zU-z+5T)wu
zIIG83u52m7C{gZLRgEm%#}eEE^m$NLlnFH<R?vf6Rf6($kp*!Cuwp+8<2@*?-F?Y(
zt!mViKL|z{(u8p<%)(YA_nQl41x>uWRcSU|s9&r&aUDE~RS-=->UV;&tjjvXdz;QN
z!9%9~tieRS48kDum|yT)q#lKC&Qpgjo8ohrYatqA!@I>Xc3qW6Mw?VT2k2jt5zOg^
zi^n{Nj{d`QBPN^Txsua*Hm1nP+=vxk;kK`}In1t#-RLomfHpw`=EiJh0dEn_TQ;&8
zGnoa91&pj2%VwmoBh4b)`t}<{V9;YCLf5#}|0g1&fCZ-}B7wE@IwI1!s6kfD!V-v4
z^XIEqs*&^|Oiw7}{#5<8LivoPol`Q1Q-*?QEY6_7FBpNJJmNfvh9fYObjAy$GjS0G
z!q|v6lQIc)><BU%G`C@p<JJtS0p0|7w+`-{CVU75$xghD;xdYNK&T5E!I8IeAg)O*
zc^xi;gb!rH(h-UG@Rj#bVEMw4YY?4;joA4e^fGMXod<ZfJp!zB<&DBFao3uHT#djJ
zNUl6;AO`G#@-Dvxs*C&vePgN`boO0=-;e;}SZe~)bqfOi7<F`_Ys@HL@+40Me9SyA
z@y!MTf{zfIWQwM_I@1qj7N;_fcH!FuN6M+5W)@2g_Zkkg!>OAbpNsPD^i#GM^9_te
Ra~B*6{Cv;9;;;DY{{o=%+XDaq

diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..eb13a03
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,4 @@
+tensorflow
+opencv-python
+numpy
+matplotlib
diff --git a/CNN-Xray.py b/scripts/CNN-Xray.py
similarity index 100%
rename from CNN-Xray.py
rename to scripts/CNN-Xray.py
diff --git a/NN2-Xray.py b/scripts/NN2-Xray.py
similarity index 100%
rename from NN2-Xray.py
rename to scripts/NN2-Xray.py
diff --git a/NN3-Xray..py b/scripts/NN3-Xray..py
similarity index 100%
rename from NN3-Xray..py
rename to scripts/NN3-Xray..py
diff --git a/Tfrecord2.py b/scripts/Tfrecord2.py
similarity index 100%
rename from Tfrecord2.py
rename to scripts/Tfrecord2.py
diff --git a/decode_tfrecord.py b/scripts/decode_tfrecord.py
similarity index 100%
rename from decode_tfrecord.py
rename to scripts/decode_tfrecord.py
diff --git a/image_pro/setup.py b/setup.py
similarity index 69%
rename from image_pro/setup.py
rename to setup.py
index 3572ce0..58145ae 100644
--- a/image_pro/setup.py
+++ b/setup.py
@@ -2,14 +2,14 @@
 
 
 setup(
-name='image_pro',
-version='0.0.1',
-description='Image processing and save into Tfrecords',
-author='',
-author_email='hilalyamine@gmail.com/alaa.el.bouchti@gmail.com',
-license='Open',
-packages=find_packages(include=['image_pro','image_pro.*','tfRecord', 'utils_data']),
-packages_dir={'tfrecord':'tfRecord',
+    name='image_pro',
+    version='0.0.1',
+    description='Image processing and save into Tfrecords',
+    author='',
+    author_email='hilalyamine@gmail.com/alaa.el.bouchti@gmail.com',
+    license='Open',
+    packages=find_packages(include=['image_pro','image_pro.*','tfRecord', 'utils_data']),
+    packages_dir={'tfrecord':'tfRecord',
               'tfrecord.tfrecord':'tfRecord/tfrecord.py',
               'tfrecord.decode':'tfRecord/decode.py',
               'utils': 'utils_data',
@@ -23,6 +23,5 @@
               'utils.norma.norma':'utils_data/contours/norma.py',
               'utils.visualize':'utils_data/contours',
               'utils.visualize.visu':'utils_data/visualize/visu.py'            
-},
-scripts = ['parameters.py']
+    },
 )
\ No newline at end of file

From b6396dec5a3adf290a69b6539e30d46e171754db Mon Sep 17 00:00:00 2001
From: a-hilaly <hilalyamine@gmail.com>
Date: Thu, 3 May 2018 21:41:01 +0200
Subject: [PATCH 28/28] add

---
 image_pro/tfRecord/Tfrecord.py              | 47 +++------------------
 image_pro/tfRecord/decode.py                | 18 ++------
 image_pro/utils_data/contours/contours.py   |  4 +-
 image_pro/utils_data/geotransf/geotransf.py |  2 +-
 4 files changed, 11 insertions(+), 60 deletions(-)

diff --git a/image_pro/tfRecord/Tfrecord.py b/image_pro/tfRecord/Tfrecord.py
index 1451b4e..fff4917 100644
--- a/image_pro/tfRecord/Tfrecord.py
+++ b/image_pro/tfRecord/Tfrecord.py
@@ -1,24 +1,19 @@
-from random import shuffle, seed
 import os
+import cv2
+import sys
 import glob
-from image_pro.utils_data.contours.contours import draw_all_contours
 import numpy as np
-
-import cv2
-
 import tensorflow as tf
 
-import sys
+from random import shuffle, seed
+from image_pro.utils_data.contours.contours import draw_all_contours
 
 
 # This function returns the different paths and corresponding labels
 def shuffling_data(data_path, labels):
     c = list(zip(data_path, labels))
-
     shuffle(c)
-
     #NOTE: data and labeld are tuples. Care about immutability 
-
     data, labeled = zip(*c) #NOTE:ith element in data correpand to ith label elmt
 
     return data, labeled
@@ -47,40 +42,27 @@ def read_paths_treat(ur_path, test_size):
 
 
 def split_data(training_prop,test_prop, data, labels, dev_prop=0):
-
     ''' 
     - training_prop is required .numerical float < 1
     - dev_prop is optional. Numerical float < 0.3
     - test_prop is required. Numerical float < 0.3
     - data and labels are lists 
     '''
-
     #NOTE: data is a list of paths of each image
 
     training_data = data[0:int(training_prop*len(data))]
-
     training_labels = labels[0:int(training_prop*len(labels))]
 
-
-
     if dev_prop != 0:
-
         dev_data = data[int(training_prop*len(data)):int((1-dev_prop)*len(data))]
-
         dev_labels = labels[int(training_prop*len(labels)):int((1-dev_prop)*len(labels))]
-
         test_data = data[int((1-dev_prop)*len(data)):]
-
         test_labels = labels[int((1-dev_prop)*len(labels)):]
-
         return training_data,training_labels, dev_data, dev_labels, test_data, test_labels
 
     else:
-
         test_data = data[int((1-test_prop)*len(data)):]
-
         test_labels = labels[int((1-test_prop)*len(labels)):]
-
         return training_data, training_labels, test_data, test_labels
 
 def load(ur_path, option):
@@ -99,14 +81,9 @@ def load(ur_path, option):
 def load_image(data, s_width, s_height, option):
 
     #cv2 load data from data (value of one path) 
-
     img_load = load(data, UR_OPTION)
-
-   
     '''img_after = cv2.cvtColor(img_load, cv2.COLOR_BGR2GRAY)'''
-    
     img= cv2.resize(img_cont, dsize=(s_width, s_height), interpolation=cv2.INTER_CUBIC)
-
     img_f = img.tostring()
     
     return img_f
@@ -127,32 +104,25 @@ def _bytes_feature(value):
 def get_tfrecord_file(ur_path, test_size, s_width, s_height,option, *x_filename,):
 
     #NOTE:data is a tuple (list) of paths of each image
-
     #NOTE: data can come from the dev-set or test-set
-
     #NOTE: Same for labeled
 
     test_img, test_label, train_img, train_label = read_paths_treat(ur_path, test_size)
-
     file_path = x_filename[0]
-
     #Open a TFRecordWriter
 
     writer = tf.python_io.TFRecordWriter(file_path)
     for i in range(len(train_img)):
         new_img = load_image(train_img[i], s_width, s_height,option)
-
         new_label = train_label[i]
-
         new_feature = { 'height': _int64_feature(s_height),
                         'width' : _int64_feature(s_width),
                         'label': _int64_feature(new_label),
                         'image': _bytes_feature(tf.compat.as_bytes(new_img))
         }
-        #NOTE: PLease refer to this url for defenirtion: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
 
+        #NOTE: PLease refer to this url for defenirtion: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
         example = tf.train.Example(features=tf.train.Features(feature=new_feature))
-
         writer.write(example.SerializeToString())
 
     writer.close()
@@ -163,24 +133,19 @@ def get_tfrecord_file(ur_path, test_size, s_width, s_height,option, *x_filename,
     for j in range(len(test_img)):
 
         new_img = load_image(test_img[j], s_width, s_height,option)
-
         new_label = test_label[j]
-
         new_feature = { 'height': _int64_feature(s_height),
                         'width' : _int64_feature(s_width),
                         'label': _int64_feature(new_label),
                         'image': _bytes_feature(tf.compat.as_bytes(new_img))
         }
-        #NOTE: PLease refer to this url for def: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
 
+        #NOTE: PLease refer to this url for def: https://github.com/tensorflow/tensorflow/blob/r1.7/tensorflow/core/example/example.proto
         example = tf.train.Example(features=tf.train.Features(feature=new_feature))
-
         writer.write(example.SerializeToString())
 
     writer.close()
-
     sys.stdout.flush()
-
     return 0
 
 
diff --git a/image_pro/tfRecord/decode.py b/image_pro/tfRecord/decode.py
index d2a5546..13e113a 100644
--- a/image_pro/tfRecord/decode.py
+++ b/image_pro/tfRecord/decode.py
@@ -1,14 +1,9 @@
-#NOTE: refer to this url: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/how_tos/reading_data/fully_connected_reader.py
-import tensorflow as tf
-import numpy as np
 import os
+import numpy as np
+import tensorflow as tf
 import matplotlib.pyplot as plt
 
-
-
-
-
-
+#NOTE: refer to this url: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/how_tos/reading_data/fully_connected_reader.py
 
 def tfrec_data_input_fn(filenames, num_epochs=1, batch_size=16, shuffle=False):
     
@@ -83,16 +78,12 @@ def _parse_record(tf_record):
 
             image_raw = tf.decode_raw(record['image'], tf.uint8)
             
-            
             def _normalize(image):
                 """Convert `image` from [0, 255] -> [-0.5, 0.5] floats."""
                 image = tf.cast(image, tf.float32) * (1. / 255)-0.5
                 return image
 
             image_n = _normalize(image_raw)
-            
-            
-
             return { 'image': image_n }
         
         # For TF dataset blog post, see https://developers.googleblog.com/2017/09/introducing-tensorflow-datasets.html
@@ -110,9 +101,6 @@ def _normalize(image):
     return _input_fn
 
 
-
-
-
 '''read_file("./tftest.tfrecords")
 
 tfrec_dev_input_fn = tfrec_data_input_fn(["tftest.tfrecords"])
diff --git a/image_pro/utils_data/contours/contours.py b/image_pro/utils_data/contours/contours.py
index 0b66f29..e503be6 100644
--- a/image_pro/utils_data/contours/contours.py
+++ b/image_pro/utils_data/contours/contours.py
@@ -15,7 +15,6 @@ def canny_cont(frame, threch1, threch2):
 ###################
 # Thresholding
 ####################
-
 def thresh_bin(frame, value_thr, value_aft):
     _ , img  = cv2.threshold(frame, value_thr, value_aft, cv2.THRESH_BINARY)
     return img
@@ -28,7 +27,7 @@ def find_contours(frame, approx=False):
         image, contours, hierarchy = cv2.findContours(tresh, cv2.RETR_TREE, cv2.CHAIN_APPROX_NONE)
 
     return image, contours, hierarchy
-    
+
     #image: source image
     # contours: contour retrieval mode
     # hierarchy: contour approximation method
@@ -123,6 +122,5 @@ def mask_pix(frame_gray):
     pixelpoints = cv2.findNonZero(mask)
     return pixelpoints
 
-
 #Aspect ratio: It is the ratio of width to height of bounding rect of the object
 
diff --git a/image_pro/utils_data/geotransf/geotransf.py b/image_pro/utils_data/geotransf/geotransf.py
index 019618d..9c9b2fb 100644
--- a/image_pro/utils_data/geotransf/geotransf.py
+++ b/image_pro/utils_data/geotransf/geotransf.py
@@ -1,11 +1,11 @@
 import cv2
 import numpy as np
 import matplotlib.pyplot as plt
+
 #Let's consider M the translation matrix. Given a translation vector (tx,ty),
 # the matrix is defined as follows: M = [1 0 tx]
 #                                       [0 1 ty]
 
-
 ##################
 # Translation
 ##################