Design And Implementation Of Hd Video Surveillance System Based On Powerpc

High-definition and intelligent are the most important parts of the development of Video Surveillance System, and it is the standard to measure the system. The higher of the video has the resolution, the clearer it plays. Intelligent depends on the later video processing. This article improved the video resolution in system architecture design and implementation, analyzed and optimized some algorithm for later video processing, improved the intelligent of whole Video Surveillance System.At present, the mainly technical of hindering video resolution is video compression and transmission, not the video collection and storage. For the design of system architecture, this article is mainly from the perspective of improve the performance of video and obtain evidence, using static image compression algorithm JPEG2000. The video compression algorithm produces large data rate, so it need some special interface to deal with. MPC8315, a kind of PowerPC family, has an eLBC, its UPM transmission style is larger data rate than PCI style, which can make up the deficiency caused by compression algorithms. The system can achieve a resolution of 1280×1024, collecting 30 frames per second, improve the resolution and achieved the high-definition of this stage for current D1 standards.The intelligent improvement of system mainly reflected in the post-video processing. This article illustrates the progress of intelligence with vehicle detection, because it was recognized as the core issues for the intelligent transportation system. High-definition of video can provide valuable information on the details, but several times the growth of data, making the current video image processing algorithms produce great delays, losing real-time requirement. This article optimizes the Gaussian Mixture background modeling algorithm through in-depth study of it, improving the speed of updates, and almost does not change the accuracy. In addition, this paper proposes the combination of aim growth and regional mark of moving object segmentation algorithm, experimental results show that the algorithm is very good.