Pest Analysis of Traffic Peccancy Forensics Service in China
Jinwei Yu, Xiying Li, Youting Zhao · 2011
Jaywalk is the most common urban traffic violation, so an intelligent jaywalker detection method is presented in this paper. The method analyses and detects jaywalker's abnormal behavior against traffic rules. Firstly, the motion object is segmented based on the background difference method, and pedestrian is recognized from vehicle according the height-width ratio. Secondly, the pedestrian's motion path is obtained through the linear predictive tracking algorithm. Then the pedestrian motion path is simplified to the line set through linear fitting, and the motion direction is calculated. Finally, the jaywalker is detected according the angle between the pedestrian's motion direction and road direction. The experiment shows that the method can detect the jaywalker effectively in short time, and it is suitable to use in real time monitoring.