With the accelerating urbanization process, road intersections, as critical nodes in transportation networks, face increasing traffic pressure due to pedestrian and vehicle violations, making order management particularly crucial. This study proposes a computer vision-based red-light violation detection system for urban intersections. Featuring a modular architecture, the system incorporates the YOLOv5 deep learning algorithm and OpenCV technology to achieve real-time video monitoring. Capable of operating in complex traffic scenarios, it demonstrates rapid and accurate identification of pedestrians and vehicles, enabling timely detection and processing of red-light running violations. Experimental results demonstrate a 95% accuracy rate in recognizing pedestrian and vehicle violations. The implemented system enhances traffic safety at urban intersections while optimizing traffic flow management efficiency.
Computer Vision-Based Traffic Monitoring: Design of a Red-Light Violation Detection System for Urban Intersections
2025-05-16
2217053 byte
Conference paper
Electronic Resource
English
Traffic violation behavior detection method based on computer vision
European Patent Office | 2021
|A Survey of Vision-Based Traffic Monitoring of Road Intersections
Online Contents | 2016
|A Survey of Vision-Based Traffic Monitoring of Road Intersections
Online Contents | 2016
|