Research And Implemenatation Of Video Storage And Retrieval And Bandwidth Adjustment Algorithm Based On H.264

As an important part of modern security industry, video surveillance products with rich and accurate information are widely used in all areas of life and work. Network video surveillance has become one of the hot spots of research. In video surveillance system, the real-time data compression, storage, retrieval, and the rational use of limited bandwidth are the key technologies. This paper studies the storage and retrieval of video data, meanwhile, the paper researches the algorithms of bandwidth-adjustment with the limited bandwidth to meet the demand of image quality.This paper researches embedded operating system and embedded database, and studies H.264 coding standard. The paper sets multi-index structure including disk index, file index, segment index and frame index for video data according to an actual project. Video data can be located precisely in the data file according to the offset information in frame index, to meet the need of accurate and rapid retrieval of video data.Based on the analysis of several factors impacting the quality of video images, the paper studies bandwidth-adjustment algorithms. The paper designs two algorithms including adjusting bandwidth averagely and adjusting bandwidth emphatically. Bandwidth is allocated reasonably according to the differences of monitor images while multi-channel videos work at the same time. These algorithms overcome the problems of mosaic and stagnant of video stream while every channel holds the same bandwidth, and meet the quality requirement of video images and the flow visibility of video storage data.In light of the actual project, the multi-index and bandwidth-adjustment algorithms are applied to the design of video surveillance system. Experiment results show that the four-stage index of video data satisfies the requirement of rapid and precise retrieval of record data, meanwhile, with the limited 1M bandwidth, the quality of video image is fine while there are mobile objects in a certain channel or some channels in multi-channel video surveillance.