Research On Motion Estimation Algorithm In Video Compression Based On H.263

With the rapid development and wide application of the multi-media communication including videophone, video monitoring and digital TV etc, the video image must be compressed in order to transmit the video. The aim of the compression is to describe the image sequence in more less bits in order to transmit in the channel. Motion estimation is the key affecting the rate of compression, and it's the focus of the video compression research.Firstly, this dissertation introduces the basic principles and method of H.263 video compression. The dissertation focuses on the FS, DS, MVFAST and Predictive MVFAST algorithms which are the typical block matching motion estimation algorithms. Basic method is summarized through the simulation and data analysis. A modified algorithms based on the MVFAST is proposed to improve the division of the motion intensity. The modified algorithm divides the big block according to the motion type and chooses search strategy and efficient termination criterion adaptively. The experiment results show that the search points of the modified algorithm decreases significantly and improve the search speed while the video quality is similar to the MVFAST and PMVFAST algorithm. At last, the thesis aimed at H.263 coder, studied some optimization techniques for improving the efficiency of motion estimation algorithm, and summarizes classical optimized schemes. The experiment results show the effectiveness of the optimized schemes.