Research Of Motion Estimation Based On H.264 Video Coding

With the rapid development of multimedia technology and Internet, the video, as one of the information core carriers, becomes the most widely in application as the reason of the most intuitive information display like video meeting and video site etc. It becomes very important to do video compression due to huge amount of video information. People propose the improved video compression standard constantly, from the original MPEG-1 to the latest H.264. As the experiments shown, H.264 can reduce 50% rate compared with H.263 and MPEG-4 under the same quality reconstructed image but it increases largely computation.As the core technology of video coding, motion estimation occupies a large number of computing time and increases the computational complexity. As the statistics shown, motion estimation occupies 60%-80% of the whole video coding time besides more reference frames, more time will be occupied. So it is very necessary to research motion estimation algorithm in the process of video compression.Firstly, the issue background and each video coding standard are introduced. Then it analyzes the H.264 basic concepts, coding principle and the key coding technology. Then it mainly focuses on the basic principle of motion estimation based on block matching and several classic fast motion estimation algorithms. In addition, make a testing and data comparison on JM10.1 platform. Finally, it proposes an improved UMHexagons algorithm which can reduce about 10% search time in the premise of ensuring the image quality compared with the current UMHexagons algorithm.