Implementation Of Moving Target Detection System Based On Zedboard

With the acceleration of China’s urbanization process,people put forward higher and higher requirements for safe cities,safe traffic and intelligent driving.As the "eyes" of the city,video surveillance technology has attracted more and more attention.The proliferation of video data makes the data storage and calculation of images more complicated.The method of using the software for image processing alone has a slow processing speed and poor real-time performance,and already has been unable to meet the increasing demand for video services.Therefore,based on the low power consumption and fast computing speed of the hardware platform Zedboard,this paper constructs a Zedboard-based moving target detection system by means of software and hardware co-processing,aiming at the moving target detection problem in video surveillance.The system uses the camera MT9V034 to capture video images and store the data in multiple VDMAs,and then sends them into the image detection module which extracts the contour of the moving target by frame difference method,and finally stores the result in another VDMA for data output display and analog alarm.The Zedboard is a practical development board based on Zynq-7000 with FPGA+ARM architecture.It can be divided into Programmable Logic(PL)and Processing System(PS).In the part of PL,the design mainly uses Vivado HLS to implement the IP core of the detection algorithm,and the two-frame difference method and the three-frame difference method are improved by adding optimization instructions and image processing methods;Then the IP core is created and called on Vivado,and the hardware platform is flexibly built to complete the integration of the detection system.In the part of PS,the data is stored and recalled by the SDK,and the analog alarm mechanism is designed to improve the security performance of the detection system.Through the test,it is found that the system has fast real-time detection speed,low system power consumption,and intuitive detection results.In addition,the system is less affected by illumination changes and other factors,so it can be used in the field of video surveillance to achieve the detection of moving targets.