Design Of Moving Vehicle Detection And Tracking System Based On Zynq

Intelligent Transportation System(ITS)is the subject of research and extensive attention in the field of transportation.The research and development of moving vehicle detection and tracking system is an important research content of ITS.Therefore,it is a great significance to study the detection and tracking system of moving vehicles with small volume,low power consumption and fast speed.The system uses Xilinx Zynq-7000 series of fully programmable system-on-chip platform,integrated on-chip dual-core ARM Cortex-A9 processor and Xilinx 7 series FPGA,using hardware and software co-design method,Image acquisition,vehicle detection and tracking,display set to an embedded system.Vehicle detection,use Xilinx advanced development tool Vivado HLS(High-Level Synthesis)to complete the image preprocessing algorithm,and use this tool to improve the three-frame difference method of moving target detection algorithm successfully transplanted to the FPGA.In the vehicle tracking algorithm,the Kalman filter motion model is established by using the centroid,size and velocity of the moving vehicle,and the Kalman filter motion model is established to predict the position of the vehicle in the next frame,and the search range is determined,and the matching based on the euclidean distance algorithm to complete the tracking of sports vehicles.The main work accomplished in this design are as follows:(1)The hardware accelerated IP core of the modified three-frame difference method is designed by high-level synthesis tool Vivado HLS,including histogram equalization,Gaussian filtering,noise removal,Sobel edge detection,morphological processing and frame difference image preprocessing algorithm.(2)Building hardware projects in graphical interface way by Vivado,including ZYNQ7 Processing System,AXI Interconnection,VDMA,HDMI and the hardware accelerate IP core and other modules,etc.(3)Building embedded Linux system on the ZedBoard development board,mainly including the Linux system transplantation,cross-compiler environment and FFmpeg +OpenCV library and Qt library transplantation,etc.(4)Realizing the system software design In the ARM part,including the driver design of VDMA IP core and the hardware acceleration IP core,the implementation of Kalman filter tracking algorithm,and building the system interface design by Qt,etc.This design adopts the method of hardware and software co-design to realize the moving vehicle detection and tracking based on Zynq,achieving the expected effect,the system has the characteristics of good real-time performance,good tracking effect and friendly interface,etc.