Anomaly Detection on Railway Track and Safety Monitoring Using Deep Learning

Railway safety is a critical concern as accidents may occur due to obstacles, human intrusion, animals, or vehicles present on railway tracks. Traditional monitoring methods mainly rely on manual inspection, which is time-consuming and not suitable for continuous real-time monitoring. With the advancement of artificial intelligence and computer vision, automated systems can analyze video data and detect potential hazards on railway tracks more efficiently. This paper presents a deep learning–based anomaly detection system for railway track safety monitoring using Convolutional Neural Networks (CNN) and the YOLOv8n object detection model. The proposed system processes real-time and stored video inputs to detect objects such as humans, animals, and vehicles on railway tracks. Detected objects are highlighted using bounding boxes, and hazardous situations trigger visual alerts along with a beep sound.