Design And Implementation Of Moving Object Detection System Based On FPGA

In recent years, with intelligent monitoring getting widespread attention, the scope of application also becomes more and more popular. In the intelligent surveillance system, moving target detection plays a very important role, accurately whether right or not of the detecting, which directly affects subsequent processing such as target tracking and understanding behavior.Based on the related feather of the large data of video image processing and real-time monitoring, the author takes the advantages of the parallel processing and pipeline of FPGA, design the system of moving object detection based on FPGA.The paper mainly describes the design ideology and implementation process of system, the related research contents are as follows:(1) In the aspect of hardware design, the paper analysises several commonly uses embedded video processing platform,then describes a design scheme which is a video processing system with FPGA as the core.The paper elaborates the overall architecture of system, the design thought of each circuit module, including the minimum FPGA system, video capture module, SDRAM cache module, video display module and so on in detail, introduces the matter of PCB board drawing need to attention, and presents the hardware physical diagram.(2) In the aspect of image processing algorithm, the paper analysises and discusses the principle and the advantages and disadvantages of three kinds of motion detection algorithm,which are the optical flow method, frame difference method and background difference method, selects the frame difference method as detection algorithm of the design, detailedly introduces the median filter and edge detection in the image preprocessing, the corrosion and expansion in the image post-processing, presents a algorithm of frame difference with edge detection.(3) In the aspect of software realization, the paper adopts a top-down design method to divide module, then detailedly discussed the realization of each module, including configuration of the register for the video decoder chip by IIC bus, decoding the digital video signal YCb Cr4:2:2, designing the logic of SDRAM controller, the timing of VGA display, the image processing algorithm, at the same time simulation-testing the relevant module by the Modelsim.(4) In the aspect of system debug and test, the paper debugges the modules online by the logic analyzer Sign Tap II, testes the practical effect of grayscale image display, median filtering, edge detection and frame difference. The test results show: each module can work normally and stablely, achieves the detection of moving targets attained the desired research purpose.