kornerFlag is an automated video assistant referee system that uses computer vision and deep learning to analyze football matches in real-time. The system detects offsides, identifies fouls, and analyzes penalty incidents with professional-grade accuracy.
| Module | Description |
|---|---|
| Object Detection | YOLOv8-powered player and ball tracking |
| Team Classification | Automatic team assignment via jersey color analysis |
| Field Homography | Camera-to-pitch coordinate mapping |
| Offside Detection | Millimeter-precise positioning at moment of pass |
| Foul Detection | Contact analysis with severity classification |
| VAR Declaration | Structured decision output with confidence scores |
- 95%+ detection accuracy
- 60 FPS video processing
- <30ms decision latency
- 8 integrated AI modules
Premier League, La Liga, Serie A, Bundesliga, Ligue 1, MLS, Eredivisie, Liga Portugal, Turkish Super Lig, and football organizations worldwide.
git clone https://github.com/Osmanttne/automated_var_full.git
cd automated_var_full
pip install -r requirements.txtMIT License - See LICENSE for details.
kornerFlag - Precision decisions for the beautiful game.