Research And Design Of Embedded Video Surveillance System Based On H.264 Encoding Algorithm

With the development of the overall economy of the society,the video surveillance system has developed rapidly in various industries such as home,school,electric power,and transportation.However,a large amount of video information will cause great difficulty in image transmission and storage without compression.As the most superior and widely used video coding standard,H.264 adopts many new technologies and new methods,which makes it better coding efficiency than the video coding standards released in the past.Motion estimation and rate-distortion optimization,as the core technology of video coding,consumes more than 70% of the entire encoding process.If the motion estimation and rate distortion optimization part can be improved,it will inevitably improve the coding performance of the encoder.This is the significance of studying the motion estimation algorithm and the rate distortion optimization algorithm.This paper first introduces the key techniques of H.264 predictive coding.Then it studies the theory of rate-distortion optimization,and proposes a new mode decision scheme that uses the relationship between motion vector and coding mode.Finally,the process of motion estimation is described.Related techniques,and several existing classical motion estimation algorithms are introduced in detail.Finally,the UMHexagonS algorithm is improved from three aspects: starting point prediction,search template and window division.The rate-distortion optimization algorithm before and after the improvement and the UMHexagonS algorithm are tested in the H.264 standard test model JM10.1.The test results show that the improved two algorithms can be used in the case of small signal-to-noise ratio and bit rate variation.Significantly reduce video encoding time.In order to test the actual performance of the improved coding algorithm on the embedded video system,this paper uses ARM11 and Linux to build an embedded video system.The improved coding algorithm is imported into the x264 encoder for video capture and encoding.By collecting video of different formats,resolutions and time,the memory size of the video file collected before and after the algorithm optimization is compared,and the optimized algorithm is obtained in the actual system.Video files take up less memory space and have a higher coding rate.The collected video information is stored on a personal computer equipped with Freenas.The external electronic device only needs to download and install the ES browser,and access the collected video information in the local area network through the FTP protocol.