Research On Motion Estimation In The H.264/AVC

H.264/AVC is the newest video coding standard of the ITU-T Video Coding Experts Group and the ISO/IEC Moving Picture Experts Group. By adopting a number of novel coding techniques, H.264/AVC has been enhanced compression performance and provision of a "network-friendly" video representation addressing "conversational" (video telephony) and "nonconversational" (storage, broadcast, or streaming) applications relative to existing standards.In this thesis, we mainly research on the new techniques in H.264/AVC, emphatically on two core techniques: motion estimation and inter prediction.First we discusses the key technique in video coding standard: motion estimation. By analyzing those available algorithms in following techniques: block matching criteria, the prediction of initial search point, search strategy and search pattern, we attempts to make an overview of fast block-based matching algorithms, deals with some typical algorithms and compare their performance. Then propose a novel motion estimation algorithm, or the MultiPattern-based Motion Vector Field Adaptive Search Technique (MPMVFAST), experiments show that MPMVFAST could obtain good PSNR performance as well as low computation cost.Inter prediction can improve the coding efficiency greatly. In order to improve the speed of inter prediction in H.264/AVC, according to variable block size motion estimation and motion selection, we proposes two new algorithms, A Fast Motion Estimation Based on Texture Analysis of Motion Vector (TAMV) and An Advanced Mode Selection Algorithm for Motion Estimation (AMSA). Experiments show that they could both obtain good PSNR performance and very fast speed as well as low bitrates.