Aiming at the vacancy of video detection in road detection

a side vehicle flow detection system is designed. The system is designed based on the architecture of digital signal processor（ DSP） and field programmable gate array（ FPGA）

which is used for front-end acquisition and data prepossessing. Parallel acceptance of data ensures the real-time performance of the system. With its unique ability of video image processing

DSP carries out arithmetic operation. By using the latest SYS/BIOS development mode programming

the development cycle of the system is greatly shortened; and the moving object detection algorithm based on virtual detection belt is improved for detection. Measure the traffic flow of two lanes perpendicular to the direction of the road. The experimental results show that the system uses a new lateral video detection mode

which can accurately identify single-vehicle traffic

accurately distinguish pedestrian-vehicle co-lane and dual-vehicle parallel

and accuracy achieves more than 90 %.