Study of Track Obstacle Detection System

Real-time tracking of obstacles on tracks improves railway safety operations through the Track Obstacles Detection System. The system employs deep learning models YOLO and SSD to quickly detect obstacles that include humans, animals and debris through reliable methods. That system enables accurate item detection and combined camera images and LiDAR sensor range for accurate information accurate identification. The NVIDIA Jetson Nano executes real-time operations as part of its hardware platform. The system employs Kalman filters as part of sensor fusion methods to provide an economical solution that scales across various industries which enhances operational performance and restricts track obstruction-related safety hazards.