Research On Block Matching Motion Estimation For H.264

H.264/AVC is the newest international video standard designed by both ITU-T VCEG and ISO/IEC MPEG. This standard can obtain high coding efficiency and excellent coding performance,and outperforms the MPEG-4 in the low bit rate environment especially. So it can meet the requirement of the network-video with low bandwidth and high quality. However, it also leads to higher computational complexity at the encoder mainly due to the fact that the H.264 encoder employs more complicated approaches and features to efficiently improve the coding performance.Hence,it is an essential research topic to reduce the high encoding complexity while maintaining the good coding performance.Motion estimation is one of the core techniques of video coding. Motion estimation and motion compensation can reduce the large amout of temporal redundancy that exists between frames of video sequences, which leads to high compression. The introduction of some new features in H.264 makes the computational complexity dramatically increase and this terribly influnces the real-time application of H.264. It is essential to optimize the encoding algorithm for the convenience of the application of H.264 in low-rate and real-time system. Thus, the research on looking for a motion estimation algorithm that can get an effective and accurate motion vector quickly becomes a hot topic recently. The research in this thesis focuses on the block matching motion estimation which is applied widely at present.First,this thesis introduces the H.264 standard and the key techniques of video coding in detail,such as intra-prediction,inter-prediction and so on. Then the basic theory of ME(Motion Estimation) is introduced,and the core techniques are analysed in detail. To reduce the computational complexity and increase the encoding efficiency in the condition of good image quality, the thesis makes improvements to ME from the aspects of mode decision,early-termination and matching criteria. The creative work in this thesis focuses on:(1) Optimize the mode dicision process by setting up a right threshold. Make early judgement to 16*16 and 8*8 block further on the basis of the exsiting early-stop for SKIP mode algorithm.(2) Optimize the motion estimation process using the early-termination technique. Choose the right threshold and terminate the motion estimation process by judging the current block is the PVB or not.(3) Improve the existing matching criteria according to its fault, and take the image soomthness into consideration on the basis of the original SAD criteria.At last, test the performance of new algorithm on JM86 platform and testify its advantage of encoding effect. The experimental results show that the encoding efficiency of the new algorithm is highly improved in the case of the encoding image quality is good enough compared with UMHexagonS algorithm which has been adopted by H.264.