Research On Vehicle Detection System Based On Microwave Radar And Video Fusion Technology

In the rapidly developing field of intelligent transportation,vehicle detection system with single sensor has been difficult to satisfy the industry the demand for diversity of information.The development trend of intelligent transportation is multi-sensor fusion vehicle detection system.The video detection system can quickly and accurately identify the target and display the real-time picture.The microwave radar detection system has the advantages of long detection distance and little weather influence.In this paper,the information detected by microwave radar and video is fused,and the advantages of multiple sensors are fully utilized to realize the multi-parameter extraction and the diversification of system functions.In this paper,the distance measurement,velocity measurement and Angle measurement are carried out by linear frequency modulated CW radar.The target echo is acquired by signal processing,which mainly includes if signal sampling,data preprocessing,filter design,target detection and target matching.The EDCA CFAR algorithm is designed,that is,the detection threshold is set according to the false alarm probability to delete the units with large fluctuation in the reference window,and the target is detected by CA CFAR.The target matching algorithm is designed,that is,according to the target power,the uphill target and the downhill target are matched,and then the Angle threshold is set to exclude false targets.Pulse canceller is used to eliminate clutter generated by fixed targets,trees,etc.The accuracy of frequency and phase detection is improved by using all phase fast Fourier transform and parabolic fitting method.This paper studies the moving target detection algorithm based on video,that is,the Gaussian mixture model is used to extract the moving target from the video,and the target information is obtained and coordinates are recorded through median filtering,expansion processing and contour extraction.In this paper,the spatial and temporal fusion of radar and video information is completed,the corresponding relationship between the space object and the target in the video is established,and the moving target is marked in the video and the parameters are displayed.This paper builds a microwave radar and video fusion processing system,and completes the preparation of relevant software.The angular reflector test and road vehicle test were carried out on the system.After passing the road test,the system achieved the expected target.The test results are as follows: the accuracy of target detection is 97.2%,the distance measurement error is less than 1m,the distance measurement range is 5?60m,the speed measurement error is less than 1.5m/s,and the horizontal Angle detection range is at most 21°.