Research On Monitoring Technology Of Crossing Violation Of Vehicle In Front Of Vehicle Video

In recent years,the problem of road traffic safety in China has become increasingly prominent.Traffic violations are the primary factor influencing road traffic safety.In the intelligent transportation system,the detection of traffic violations has become a focus of research at home and abroad.In particular,violations such as changing lanes and turning around can easily lead to traffic accidents.However,the current "electronic police" system has a relatively small coverage compared with the vehicle conditions.Therefore,violation detection of vehicles under vehicle video becomes the focus of traffic violation detection.This paper is based on the analysis of the research and application of video based vehicle violation monitoring technology at home and abroad.Aiming at the problem that the traditional fixed overrun behavior monitoring system has high input cost and can not realize the whole network coverage,this dissertation focuses on the automatic monitoring technology of the overrun behavior of the overriding vehicle in front of vehicle based on vehicle video.The main research contents and research results include:First,the research status and development trend of video-based vehicle detection research are summarized.Analysis based on the working principle of the Deformable Part Model(DPM)and its improved algorithm FTVQ method,combined with the need of cross-line violation behavior detection,a rapid detection method based on improved DPM is put forward,and under different scenarios of different data sets method validation;Second,analyze the characteristics of vehicles to identify the status of research at home and abroad,In this paper,vehicle license plate information is taken as the vehicle feature identifier.Based on the traditional license plate recognition process framework,in order to improve the efficiency of license plate character recognition,a fast license plate information recognition method based on Hausdorff algorithm is proposed;Finally,for the problem of violation of the double yellow line in front of the vehicle,Based on the comprehensive utilization of the geometric characteristics and image features of lane line,a method of determining the cross-line behavior based on local matching of image masks is proposed.