Research On Vehicle Tracking Technology Based On Roadside Camera

Vehicle tracking technology is one of the key technologies in the field of intelligent vehicles.It can not only reduce the incidence of traffic accidents,but also improve the efficiency of road traffic.It has important research significance.Urban road traffic environment is more complex.Vehicle tracking based on vehicle camera can not obtain all traffic environment information,It needs to cooperate with roadside equipment to obtain complete traffic environment information.LTE-V-based vehicle networking technology and networked intelligent driving technology have unique advantages in China.In order to explore the application prospects of LTE-V technology in the field of Intelligent Network Unified Vehicle,this thesis studies vehicle tracking technology based on roadside camera.Firstly,the vehicle tracking technology based on roadside unit is studied preliminarily.Vehicle detection,vehicle tracking and vehicle-road communication under the roadside unit are analyzed and compared.According to the characteristics of vehicle tracking in the roadside unit,ViBe algorithm is determined as the vehicle detection method,Kalman filtering algorithm is determined as the vehicle tracking method,and LTE-V as the vehicle-road communication.Secondly,vehicle detection method based on roadside camera is deeply studied.Firstly,the main road conditions of China are analyzed.Then,on the basis of the traditional ViBe algorithm,this algorithm is implemented,which successfully reduces the ghost and shadow phenomena in vehicle detection.The traditional vehicle speed detection method based on virtual-loop is improved,so that the vehicle motion information detection method proposed in this thesis can simultaneously detect the vehicle speed information and acceleration information.Finally,the vehicle tracking method based on roadside unit is deeply researched,and the effectiveness and superiority of the vehicle tracking method proposed in this thesis is verified by experiments.Firstly,the vehicle tracking module based on the roadside camera is designed,which Combines the improved ViBe algorithm with Kalman filter algorithm,a weak overlapping visual field tracking module is constructed,so that the vehicle can be continuously tracked by the roadside unit while driving on the road.Then,a vehicle information communication scheme based on roadside communication equipment is designed to ensure that the vehicle information obtained under the current roadside unit can be accurately transmitted to the next roadside unit.Finally,the vehicle tracking based on roadside camera proposed in this thesis is experimented,which verifies the effectiveness and superiority of the vehicle detection and vehicle tracking method in this thesis.Based on the improved ViBe algorithm,this thesis reduces the influence of ghost phenomenon and shadow foreground phenomenon on vehicle detection,and combines the improved ViBe algorithm with Kalman filter algorithm to improve the detection accuracy of moving vehicles,which has good theoretical significance.The vehicle tracking method proposed in this thesis is tested on the DMD31 module of Datang Telecom,which provides some reference value for the practical application of the vehicle tracking method proposed in this thesis.