Intelligent Video Surveillance System Research And Implementation Based On BF533

With the rapid development of computer technology and information technology, wide application in every field and all kinds of video image processing algorithm is put forward and constantly improved, video surveillance system has made a great development over the past few decades. The past video monitoring has a lot of weaknesses, such as:big volume, not easy to install, poor stability, maintenance is very troublesome, poor real-time performance, take up too much room, image processing function is single, intellectualized degree is very low. These problems have been greatly improved. Now, most of the schemes is based on embedded technology, use a lot of image processing algorithm, and can achieve detecting, tracking and behavior understanding for the moving target, It gives full play to the advantages of embedded system, the short system, low power consumption, strong real-time performance, the storage, transmission and display of image is very convenient. In this paper, the main research work done is as follows:First, three commonly used kinds of moving object detection algorithm have been deeply researched and analysised. Optical flow method, frame difference method, background difference method. Compared the advantages and disadvantages of them, choose a relatively good background difference method as the focus in this paper. Background modeling is the key to the way, so we further studied the variety of background modeling method, relatively well mixed gaussian distribution model is eventually chosen as our modeling method. And then, with the OpenCV, we realize the algorithm on VC++6.0. The experimental results showed that the algorithm can get a good effect of moving target detection.Later, after full analysis of the system requirements, we design a set of video monitoring system based on embedded processor BF533of ADI company, the new DSP chip is designed for multimedia applications by ADI company, fast operation, low power consumption, it is very suitable for videodata acquisition and processing. We have given the system’s overall design:video acquisition module, a video storage module, video processing module and video transmission module. And we have made the detailed introduction of each module.Then, we will complete a development of the software program. Firstly, introduce integrated development environment VisualDSP++5.0. In this environment, we can write, compile, download and debug software program. Because of the multitask scheduling management, the μC/OS-Ⅱ real-time operating system has been introduced. And we have introduces the system transplantation method in detail. Finally, give the overall design of software system, including:data acquisition tasks, data processing, data compression, data transfer tasks, and introduce the software process and code in detail.Experimental results show that the design of video monitoring system in this paper basic meet the system requirements, and can process the collected video image for the motion target detection, quickly and accurately extract the moving object. And the results can be rapidly transmitted to a remote PC, so that monitoring personnel can timely feedback the testing result.Finally, the full paper is summarized, and point out the shortage of this article and needs to be further improved.