Football Tracker - Computer Vision

This project is an advanced computer vision system designed for real-time football (soccer) analytics. It leverages YOLO (You Only Look Once) for object detection and Supervision for post-processing and annotation management to detect players, count them, classify their team affiliations, and track the ball’s movement. The system provides a multi-display interface, including a main frame showing the live camera feed with player detection and team classification, a bottom frame displaying a simplified top-down view of player positions, and a separate ball tracker for monitoring the ball’s movement. Built using state-of-the-art tools and algorithms, this tool is ideal for coaches, broadcasters, and fans, offering deep insights into player formations, team dynamics, and ball possession during the game. Key Features: Player Detection: Accurately detects players on the field using YOLO for real-time object detection. Player Counting: Automatically counts the number of players in real-time using Supervision for post-processing and data management. Team Classification: Classifies players by team using jersey color or pattern recognition, highlighting them with team-specific colors. Football Frame: Displays a simplified top-down view of player positions on a bottom frame for easy analysis of formations and movements. Ball Tracking: Tracks the ball’s movement in real-time using YOLO for detection and Supervision for tracking and annotation. Multi-Display Interface: Main Frame: Live camera feed with overlaid player detection, team classification, and player count. Bottom Frame: Simplified football field showing player positions. Ball Tracker: Separate display for real-time ball tracking. Tools Used: YOLO: For real-time player and ball detection. Supervision: For post-processing, tracking, and annotation management. OpenCV: For real-time rendering and visualization. Applications: Coaching, live broadcasting, and fan engagement.