Research On Block Matching Motion Estimation Algorithms In H.264

H.264 achieves very low bit rate, high quality and high compression ratio. However, its highly efficient compression is at the cost of increased computational complexity. Motion estimation occupies the most time in the process of video coding. For this reason, fast motion estimation in H.264 has always been a hot spot in the video compression field.Firstly, this thesis discusses the basic principle and the essential technologies in H.264 video coding standard. Secondly, it introduces several popularly fast block matching motion estimation algorithms and analyzes their advantages and disadvantages. Thirdly, in order to cover the shortage of the fast block matching motion estimation algorithms, this thesis proposes an efficient and fast block matching motion estimation algorithm——the Double Big Small Square(DBSS) algorithm. One strategy of this algorithm is to begin searching from two positions the predictive search center and the (0,0) position, which can still hold direction when the relatively predictive error is big. The other strategy is to adopt the Big Small Square search pattern, which chooses search step according to the moving state of objects. A series of skills, such as initial search center prediction, the diamond search algorithm and successively sampled distortion measure, are also adopted in this algorithm.Experimental results show that the DBSS algorithm can highly reduce computational complexity with little video quality dropping compared with FS. Also this algorithm is superior to UMHexagonS in terms of both speed and quality.