GR178: Athlete Vision Check

Prof Ir EUR ING Dr Vinesh Thiruchelvam APU

This project introduces an integrated system for running performance analysis and injury prediction using motion capture and AI technologies. The system combines OpenPose-based markerless motion tracking, CNN-driven gait analysis, XGBoost injury prediction, and wearable sensors for stride and speed monitoring. OpenPose extracts 2D body keypoints for CNN analysis of gait metrics, while XGBoost forecasts injury risks using training and biomechanical data. The ESP32-powered wearable system transmits real-time stride data via WiFi, visualized through a Python GUI. Achieving ≥0.73 correlation with gold-standard data, the solution offers accurate, scalable, and field-ready performance optimization for athletes and sports professionals.