04In Development
Football Turf Analytics
Real-time football analytics system using computer vision to track players, detect events, and surface insights on a live dashboard.
Overview
An ongoing computer vision project that processes live camera feeds from indoor turf venues to extract football analytics. Uses YOLO for player and ball detection, jersey color clustering for team classification, and rule-based logic for event detection — all surfaced on a real-time dashboard.
Key Features
- 01Player and ball detection using YOLO
- 02Team classification via jersey color clustering
- 03Event detection: passes, shots, goals, and fouls
- 04Ball possession tracking per team
- 05Player distance covered and heatmap generation
- 06Perspective calibration to convert pixel coordinates to real-world meters
- 07Live analytics dashboard
Technologies
Backend
Python
AI / ML
YOLOv8, OpenCV, K-Means Clustering
Tools
Perspective Calibration (pixel → meters)