Computer vision system using YOLO and Faster R-CNN enables real-time player tracking and positioning analysis for soccer coaching decisions

Researchers developed an object detection and tracking system to identify and track player positions throughout soccer matches using video footage

The system uses YOLO and Faster R-CNN models evaluated on custom video data to achieve accurate player identification

A point prediction model identifies key field markers and combines them with known field dimensions to extract actual player distances and positioning

The technology provides actionable data to inform coaching decisions, improve individual player performance, and enhance overall team strategies

The system integrates with SAM (Segment Anything Model) to optimize object identification accuracy for complex game analysis