Based On The H.264 Motion Estimation Algorithm

Video information in multimedia field occupies an important position, but the video data redundancy is usually very large. Therefore, the compression of the video directly determines the quality of the services. Next-generation H.264 video compression standard is capable of low bit rate, high-quality, high compression ratio and so on. After the analysis of H.264 encoder we can be conclude that motion estimation shares the longest processing time in H.264. Therefore, how to ensure that the premise of the performance of motion estimation algorithm and improve the efficiency of current video compression has become one of the hot areas of research.The paper first discusses the H.264 video coding standard of the basic structure and key technologies; then the basic principles of motion estimation are described. The paper gives the full review of several classic algorithms based on the block matching method. Understanding the nature of block-matching, we systematically summarize the current block matching guidelines and algorithms in China and abroad. According to different search algorithms used in the search template and the search strategy, we analyze the scope of application and their advantages and disadvantages.Secondly, we give some aspects of motion estimation to do a detailed analysis and research: analyzing two SAD sub-sampling techniques from the distribution of sampling points and sampling rate. After analyzing the existing initial search point prediction method, a vector similarity based on the initial search point prediction method is proposed; At the same time, the existing distribution model of the motion vector is introduced, and experimental results are shown. Then an adaptive search strategy based on the direction is applied, and we discuss its application and advantages based on the experimental verification.In addition, we discuss on the sub-pixel motion estimation technology and multi-reference frame selection algorithm and analyze the EPZS algorithm of H.264 standard model with JM, combining with previous research results on the improved IEPZS algorithm. Tested with a number of the typical video sequences, the search algorithm can reduce the estimating time, and accelerate the entire H.264 encoding system under an assuring performance.