The Application Research On Remote Video Surveillance System Based On H.264/AVC

In recent years, with the unceasing development of the network technology, the transmission technology and the video coding technology, the remote video surveillance system obtained the widespread application in every walk of life. Compared with the former video coding standard, because H.264/AVC has used the technologies like the multi-mode selection, rate control and so on, has a higher compression ratio and the stronger network compatibility, so it becomes a hot spot that how to apply the H.264/AVC video coding standard in the remote video surveillance system.Firstly, this paper introduces the key modules of remote video surveillance system and the new techniques that used in the standard of H.264/AVC video coding in detail .Then, in view of the high complexity of current mode decision algorithm, with the analysis of video sequence images'mode decisions in the video surveillance system, summed up the characteristic that a lot of SKIP modes and 16×16 modes that used in the frames of video surveillance images.Thus this paper presents a SKIP mode and 16×16 mode prediction algorithm. In addition, to solve the problem of too many candidate modes in coding moving macroblock, an some modes pre-exclude method is proposed by considering the much lower uses of some modes in the video surveillance system. From the two different aspects, the paper constituted a fast inter mode decision algorithm used in video surveillace system.The experimental results show that there is enormous complexity reduced and great encoding time saved compared with the original H.264 reference software. Finally, according to necessary to carry on the examination to all monitoring images in the block matching motion detection, wasting memory resources and using motion vector to exam alone has conclusion in the robustness and sensitivity , this paper also proposes a improved block matching motion detection algorithm which is based on H.264/AVC .The algorithm can save the massive nonessential motion detection process, save the surveillance system's memory resources, and can efficiently reduce the influence of brightness change and lens dithering.The results of simulation and research show that the fast inter mode decision algorithm and the improved block matching motion detection algorithm proposed in this paper are much more suitable for remote video surveillance system.