Research And Optimization Of The Block Matching Motion Estimation Algorithm Based On H.264/AVC

In the21st century, this great information era has already entered the second decade, as our Internet is becoming more and more prompter, computer multimedia technology is accelerating development, digital video related field were constantly expanded, network video service has been greatly demand, however, the data quantity of video information is very tremendous, which the network bandwidth and storage equipment caused great pressure. Therefore, it is urgent to improve video coding algorithm, that can be more efficient to solve video data transmission and storage problem. H.264/AVC is a new generation of video coding standard which are developed by Joint meeting expert group composed by ITU-T VCEG (video coding expert group) and MPEG ISO/IEC (activity image coding expert group) two international organization for standardization. It has higher coding efficiency and high quality video screen and stronger ability to adapt to the network, at the same time, adopt hybrid coding structure, and can adapt to different occasions, etc.Motion estimation is the core technology of the video compression coding, it is the main method to remove time redundant, its time consumption accounted for60%to80%of the whole coding process, motion estimation algorithm directly determine the quality of video coding efficiency, so the high quality estimation algorithm is to premise condition of making video encoding become effective. Block matching motion estimation algorithm (BMA), due to the algorithm is simple and easy to hardware implementation, that are widely used in all kinds of video coding standard. So research a kind of fast block matching motion estimation algorithm for video compression coding has important practical significance.In this paper, the h.264video compression coding standard and motion estimation of the key technology has been analysed deeply, after introduce the classical motion estimation algorithm, The research highlighted the asymmetric cross multi-level hexagonal lattice point search algorithm UMHexagons, and in the light of the shortcomings of UMHexagons algorithm put forward the improvement of the following five aspects:5×5spiral search, multi-level hexagon search, extended hexagon search the three search template optimization makes the algorithm can quickly find the best matching block, and at the same time, can greatly reduce search point; For Static Block Advance Judging can reduce a lot of unnecessary search, earlier complete matching process; Incomplete pixel error matching can shorten each pixel block calculation time, makes the calculation between block and block reduced by half.Finally, through the test JM16.2model, perform the experiment environment of algorithm validation in the VS2008platform, the experiment proved that the improved algorithm can effectively relieve calculation and maintain video quality, compared with the original UMHexagons algorithm, the new algorithm has better performance, which can reduce the average search points by45%, and can save the motion estimation time by12.69%-36.85%. that achieve the purpose of improving the coding real-time.