Advanced Driver Assistance System based on Machine Vision

Abstract

In view of the frequent occurrence of traffic problems such as urban traffic congestion and traffic accidents, people pay more and more attention to traffic safety. This paper takes the key technical problems such as front vehicle detection and identification and pre-collision detection in the advanced driving assistance system as the research object, this paper puts forward a computer vision solution based on YOLOv5 algorithm and monocular camera distance calibration, which provides more comprehensive driving environment information for the advanced driving assistance system and improves the active safety of the vehicle. Finally, a case study is given to verify the superiority of the algorithm proposed in this paper.