In the current social environment, pedestrian road safety has become the focus of widespread attention, especially in complex traffic environments, and the effectiveness of pedestrian detection technology is particularly important. The purpose of this study is to improve the accuracy of pedestrian detection and tracking by combining YOLOv8 and ByteTrack technology to meet the detection challenges in complex environments. Experimentally, the pedestrian model is trained by YOLOv8, and the ByteTrack algorithm is used to perform multi-target tracking after the optimal model is obtained, which solves the tracking failure problem of the traditional method in the case of occlusion. In the real-time judgment process of the dangerous area, the advanced and efficient Ray Casting algorithm is adopted, and the warning mechanism is triggered when a pedestrian is detected entering the dangerous area. This study significantly improves the accuracy of pedestrian detection and warning in complex environments, provides advanced technical support for traffic safety management systems, and is of great significance for the development of smart urban transportation systems.
Detection and warning of pedestrians in dangerous areas using YOLOv8 and bytetrack multi-target tracking technology
International Conference on Smart Transportation and City Engineering (STCE 2024) ; 2024 ; Chongqing, China
Proc. SPIE ; 13575
2025-04-28
Conference paper
Electronic Resource
English
DANGEROUS VEHICLE WARNING SYSTEM FOR PEDESTRIANS AND BICYCLES
European Patent Office | 2020
|DANGEROUS VEHICLE WARNING SYSTEM FOR PEDESTRIANS AND BICYCLES
European Patent Office | 2020
|