The Design And Research Of Moving Object Detection Based On Fpga

Detection moving objects in serial images is the mostly research file in computer vision and digital image processing, It has great significance for realizing the stable and fast mobbing object detection system.The subject focuses on the algorithm design and hardware implement of an embedded digital video monitoring system based on FPGA technique.The paper firstly analyses the basic principles of three main detection algorithms of moving object, and discusses advantages and disadvantages of them. In this paper, the classical algorithms of edge detecting are analyzed, based on these, the author tries to take the algorithm of edge detecting and algorithm of moving object together for detecting the moving target. After simulated,the algorithm can detect the edge of the moving object.The benefits of this algorithm are simple, small amount of calculation, strong adaptability for the dynamic environment, good robustness. The algorithm is adaptability used in the real-time processing system. If we want to get more accurate and better information on the edge of the image display, edge refinement and discrete points removal technologies should implemented. Then the system design architecture and the FPGA design board are introduced. Finally, the system is divided into some functional modules and each functional module implemented on FPGA is presented in detail.In this paper through analyzing the technologies of moving object detection, a three frame difference algorithm based on edge detection algorithm is adopted, with the task of this system; a real-time motion detection system is developed based on FPGA. Using FPGA to realize system design will improve the speed of system, it also has such advantages as flexibility and adaptability.