The Research Of Motion Estimation Based On H.264/AVC

Today is the era of information explosion,and the information throughput increased dramtically,while the video as the most important carrier of the information,the data it hosted is enormous,therefore it is necessary to compress video information,so that it can transfer in the case of the limited bandwith.H.264/AVC standard produced in such a case,it has a very high coding efficiency,while the complexity of the algorithm compared with the original standards improved greatly,and it will also increase the time of coding.In order to be effectively applied in real-time occasions,it is necessary to optimize its coding algorithm.The motion estimation as one of the core technologies of videdo coding,and the most time-consuming part of the entire coding process,which accounts for about60%of the entire coding time,it is become a hotspot to do research about the optimization of the mostion estimation algorithm.This paper firstly describes the research background and the development process of the video compression standards,then introduces the basic concepts related to the standard,the structure of the codecs,and make detailed analysis for the key technologies of coding.And then focus on the research of estimation algorithm based on block matching,including block-matching principle,several classic block-matching algorithm,and analyze their strengths and weakness.On the basis,make a depth analysis for the UMHexagonS algorithm and propose optimization of twp aspects:the first is the classification of the intensity of motion whitch reduces the choices of the split mode;the second is the extension of the hexagonal search template in UMHexagonS algorithm which improved the search accuracy. Finally in the test model of JM10.1,make a data comparison analysis between the UMHexagonS algorithm and the Improved UMHexagonS algorithm.Then make a conclusion:in the case of the bit rate and signal-to-nosie ratio is almost the same,the improved UMHexagonS algorithm in motion estimation time reduces on average of24.9%compared to the UMHexagonS algorithm,and the total encoding time reduces on average of12.2%.This verify the effectiveness of the improved algorithm.