Football Analytics Pipeline

This project is a football analytics pipeline that takes a video of a football match as input and outputs a video with various analytics features, including player tracking, team assignment, ball possession, speed and distance estimation, and camera movement estimation.

football.mp4

Features

• Player Tracking: Tracks the movement of each player on the field.
• Team Assignment: Assigns players to their respective teams based on jersey color.
• Ball Possession: Determines which player is in possession of the ball.
• Speed and Distance Estimation: Calculates the speed and distance covered by each player.
• Camera Movement Estimation: Estimates the movement of the camera during the match.
• View Transformation: Transforms the perspective of the video to a top-down view.

Requirements

• Python 3.7+
• OpenCV
• Ultralytics YOLO
• NumPy

Installation

1. Clone the repository:

   git https://github.com/ROBERT-ADDO-ASANTE-DARKO/Football-Analysis-with-YOLOv8-and-OpenCV.git

2. Navigate to the project directory:

   cd Football-Analysis-with-YOLOv8-and-OpenCV

3. Install the required packages:

   pip install -r requirements.txt

Usage

1. Place the video file you want to analyze in the input_video directory.
2. Run the main script:

   python main.py

3. The output video will be saved in the output_videos directory.

File Structure

Football-Analysis-with-YOLOv5-and-OpenCV/
├── main.py                      # Main script
├── utils.py                     # Utility functions
├── trackers.py                  # Object tracking module
├── team_assigner.py             # Team assignment module
├── player_ball_assigner.py       # Ball possession assignment module
├── camera_movement_estimator.py  # Camera movement estimation module
├── view_transformer.py           # View transformation module
├── speed_and_distance_estimator.py # Speed and distance estimation module
├── input_video/                 # Input video files
├── output_videos/               # Output video files
├── models/                      # Trained YOLO model
└── stubs/                       # Stub files for intermediate results