Lane Detection Project

This repository contains scripts and notebooks for lane detection using OpenCV and NumPy. The project demonstrates lane detection on both video streams and single images, providing step-by-step image processing pipelines and visualizations.

Contents

• lanedetection_video.py / lanedetection_video.ipynb: Lane detection on video files.
• lanedetect_single_image.py / lanedetection_single_image.ipynb: Lane detection on a single image.
• images/: Sample images and intermediate processing results.
• video/: Sample input and output videos.
• requirements.txt: Python dependencies.

---

Lane Detection on Video

This script performs lane detection on a video stream. Each frame is processed to detect and highlight lane lines, providing a real-time visualization.

Video Script Overview

• Input: A video file
  
• Output: Video frames with detected lane lines overlaid, displayed in a window
  

Video Script Main Steps

1. Video Loading: Opens the video file and processes it frame by frame.
2. Image Snipping: Crops the region of interest from each frame to focus on the road area.
3. Masking: Applies a polygonal mask to isolate the lane area.
4. Thresholding: Highlights white and yellow lane markings using color masks in HSV color space.
5. Blurring: Applies Gaussian blur to reduce noise and improve edge detection.
6. Edge Detection: Uses the Canny algorithm to find edges in the frame.
7. Line Detection: Uses the Hough Transform to detect lines corresponding to lane markers.
8. Line Averaging and Drawing: Averages detected lines using a moving average for stability, then draws the left and right lane lines on the frame.
9. Display: Shows the processed frame with detected lanes in a window. Press 'q' to quit early.

Video Script Functions

• snip_image(image): Crops the lower part of the frame to focus on the road.
• mask_image(image): Applies a polygonal mask to the frame to isolate the lane area.
• thresh_image(image): Applies color and grayscale thresholding to highlight lane lines.
• blur_image(image): Applies Gaussian blur to the frame.
• edge_image(image): Detects edges using the Canny algorithm.
• lined_image(image): Detects lines using the Hough Transform.
• show_line(lines, snip, color): Draws averaged lane lines on the frame using a moving average for stability.
• moving_average(snip, ptsL, ptsR): Maintains a moving average of detected lane line points for smoother visualization.
• main(): Orchestrates the lane detection pipeline for each video frame and displays the result.

How to Run the Video Script

Run the script directly:

python lanedetection_video.py

A window will display the video with detected lane lines. Press 'q' to exit before the video ends.

---

Lane Detection on a Single Image

This script performs lane detection on a single image. The process involves several image processing steps to isolate and highlight lane lines in a road image.

Single Image Script Overview

• Input: A road image
  
• Output: The input image with detected lane lines overlaid
  

Single Image Script Main Steps

1. Image Snipping: Crops the region of interest from the bottom of the image where lanes are expected.
   
2. Masking: Applies a polygonal mask to focus on the lane area.
   
3. Thresholding: Converts the image to HSV and grayscale, then applies a mask to highlight white lane markings.
   
4. Blurring: Uses Gaussian blur to reduce noise.
   
5. Edge Detection: Applies Canny edge detection to find lane edges.
   
6. Line Detection: Uses Hough Transform to detect lines corresponding to lane markers.
7. Drawing Lines: Averages detected lines and draws them on the original snipped image.
   

Single Image Script Functions

• snip_image(image): Crops the lower part of the image to focus on the road.
• mask_image(image): Applies a polygonal mask to the image to isolate the lane area.
• thresh_image(image): Applies color and grayscale thresholding to highlight lane lines.
• blur_image(image): Applies Gaussian blur to the image.
• edge_image(image): Detects edges using the Canny algorithm.
• lined_image(image): Detects lines using the Hough Transform.
• main(): Orchestrates the lane detection pipeline and displays the result.

How to Run the Single Image Script

Run the script directly:

python lanedetect_single_image.py

The result will be displayed in a window showing the detected lane lines.

---

Requirements

• Python 3.x
• OpenCV (cv2)
• NumPy
• imutils

Install dependencies with:

pip install -r requirements.txt

Notes

• The scripts are designed for demonstration and may require parameter tuning for different images, videos, or camera perspectives.
• The region of interest and mask polygon are hardcoded for the sample data and may need adjustment for other inputs.
• The video script uses a moving average to stabilize lane line detection across frames.

---

Author: Ian Edmundson
Date: August 2025