VisionPlay

This project aims to use YOLO and several computer vision techniques to detect objects, segment players, measure their speed, and analyze the scene's perspective. The following methods are employed:

1. YOLO Object Detection: Used to detect players in images and videos.
2. K-Means Clustering: Applied to segment players from the background and identify their t-shirt color.
3. Optical Flow: Used to measure camera movement and track object motion.
4. Perspective Transformation: Utilizes OpenCV's perspective transform to represent the scene's depth and perspective.
5. Speed and Distance Measurement: Calculates the player's speed and distance covered based on the detected movement.

Requirements

To run this project, make sure you have the following libraries installed:

• Python 3.x
• YOLOv (Ultralytics)
• OpenCV
• NumPy
• Scikit-learn

Installation

Clone the repository

git clone https://github.com/ahmedanwar123/VisionPlay.git

cd VisionPlay

Using pip/venv

Create virtual environment

python -m venv football

Activate the environment

For Unix/MacOS:

source football/bin/activate

For Windows:

football\Scripts\activate

Install requirements pip install -r requirements.txt

Using Conda

Create the conda environment

conda env create -f conda-env.yaml

Activate the environment

conda activate football

Features (Subject to edits)

• Object Detection: Uses YOLO to detect players and other objects in real-time from image or video input.
• Player Segmentation: Segments players from the background using K-Means clustering and identifies t-shirt color.
• Camera Movement Tracking: Measures camera movement through optical flow analysis.
• Scene Perspective: Transforms the scene's perspective to represent depth using OpenCV.
• Speed & Distance Calculation: Measures the player's speed and distance covered on the field based on detected motion.