Research On Key Equipment For Real-time Collection Of Urban Traffic Flow Based On Video Collection Technology

With the rapid development of economy,the increasing number of private cars has led to the saturation of urban road capacity,and the traffic congestion problem has become an urgent problem to be solved.To solve the problem of traffic congestion and complete the coordination of road traffic,the most important link is to collect traffic information.There are many ways to collect real-time traffic information data,and the image processing technology which has developed rapidly in recent years has been gradually applied to traffic flow information collection.It has the advantages of large area detection,flexible setting and easy updating,and has now become a research hotspot of intelligent transportation.This paper designs and implements a kind of key equipment for real-time collection of urban traffic flow based on video collection technology.It collects traffic video images of urban roads by installing camera equipment at road junctions and sections.The collectedvideo is processed in real time through embedded equipment.The method of combining embedded technology with digital image processing technology is adopted to process various traffic data from the collected video flow.Including traffic flow,average vehicle speed,congestion status,lane occupancy rate and headway,etc.,and these data are formed into data packets according to a communication protocol and sent to a remote server for subsequent traffic flow big data processing and overall control by traffic managers.This paper realizes the real-time acquisition of video stream on OK6410 embedded development equipment.the camera and OK6410 are connected by USB.the encoding and decoding method of video stream adopts H.264 standard.in the aspect of transmission,the open source code project luvcview is used to complete the acquisition of images encoded by the camera.The acquisition back-end uses OK6410 embedded development equipment based on ARM11-bit architecture,completing the compilation of four driver modules based on embedded Linux system in video acquisition,including USB interface,LCD,touch screen and DMA controller.The main processing program is completed in the Linux user layer.OpenCV2.6.9 version 2.6.9 is configured in the embedded Linux3.0.1 system environment to realize the development of image processing program.The image processing algorithm optimizes the image enhancement,improves the processing speed,and meets the requirements of real-time system data acquisition.The Qt4.4.3 version and the tslib library are configured in the embedded device,and the graphical interface of the development board and the LCD touch function are designed and implemented.The processed road segment data is connected to the RS232 serial interface of the embedded system through a wireless digital transmission radio station to transmit the data packet to a subsequent server,and in the transmission process,a TDMA-based transmission protocol mode is used,and acquisition nodes of a plurality of road segments are divided into different time slots to transmit the data packet.The test results show that the detection system has fast real-time,accuracy and reliability for the information collection of vehicles at intersections,and the transmission protocol test for the collected data to the server also meets the design requirements,facilitating the subsequent data processing and prediction based on traffic data.