Research And Implementation Of Intelligent Video Surveillance System In Fixed Background

With the video surveillance system widely used, deficiencies of traditional video surveillance system appear gradually, and intelligence of video surveillance system has been an inevitable development trend. Aiming at the shortcoming of traditional video surveillance systems, oriented to application environment of the fixed camera and low-motion scene, this paper presents an integrated design and solution of intelligent video surveillance system suitable for fixed background, studies deeply some key technologies of intelligent video surveillance system, and implements the archetypal system.Motion detection is the basis of intelligent video surveillance system. With the in-depth analysis and research of a lot of motion detection algorithms, this thesis chooses frame-difference method as the system's motion detection method. In view of the disadvantages of large amount of calculation and incomplete extraction of moving object of frame-difference method based on pixel, this paper brings forward an new frame-difference method based on a 16 *16-pixel block. The method not only reduces the calculating amount, thus ensure the real-time of motion detection, but also achieves the complete extraction of moving objects.On the basis of inter-frame motion detection, this paper improves the imperfection of not removing redundant information between frames and low video compression ratio of MJPEG video compression algorithm. The improved compression algorithm only deals with moving objects area which is extracted by motion detection, and ignores unchanged background, so video compression ratio is improved greatly while ensuring the quality of video. The pressure of storage and network transmission of video data is decreased.In order to ensure real-time of systems, this article also optimizes design of video capture module and realization of video compression algorithm. Video capture module uses Ping-pong buffer structure which allows the video capture and processing to perform simultaneously, also realizes the sub-sampling and separate storage of YUV video data which improves video compression ratio and provides convenience for following video processing. According to DSP architecture features, this paper implements the optimization of video compress algorithm on DSP, which makes the speed of video compression improved greatly.