Detection Of Railway Accident Risk Using Deep Learning Approach

Railroad safety is given top attention in modern transportation networks, with a focus on preventing accidents and preserving passenger safety. Using state-of-the-art deep learning algorithm YOLOv8 and computer vision techniques, this study presents a novel Railway Passenger Safety System (RPSS) that can detect and predict potentially hazardous passenger behaviors near level crossings and track crossings. The RPSS records visual data in real-time from CCTV camera footage and enables continuous railroad environment monitoring. YOLOv8, renowned for its outstanding object identification skills, is utilized to identify various passenger activities, such as illegal crossings, improper proximity to tracks, or other potentially harmful actions. The system may promptly notify authorities and train operators when risky acts are detected, facilitating prompt action to avert mishaps. This study discusses the RPSS’s architecture and design, including model training and data gathering. The system’s effectiveness is evaluated using a huge dataset of passenger activities, and the results demonstrate that it can swiftly and reliably recognize and forecast dangerous postures. The Railway Passenger Safety System, which is the subject of this research, represents a noteworthy advancement in railway safety technology. It provides a proactive approach to enhance passenger safety, reduce accident rates, and augment overall railway safety. YOLOv8 and computer vision in the rail sector could revolutionize safety protocols and increase everyone’s safety when traveling by train.