The Research Of Fast Intra-prediction Algorithm In H.264

ITU-T H.264/MPEG-4 Part 10 AVC (H.264/AVC for short) , which wasestablished by Joint Video Team (JVT) consisting of experts from ITU-T's VideoCoding Experts Group (VCEG) and ISO/IEC's Moving Picture Experts Group(MPEG),is the newest international video coding standard, and contains manyadvanced video coding technologies and thoughts. Compared to those existing videocoding standards, H.264 has made a big progress in coding efficiency, and a friendlyinterface of network and fault tolerance. But the advanced coding compressionefficiency of H.264 make the computational complexity have a dramatic increase ,compared to the simple profile of MPEG-4, its coding complexity have increased by10 times, so it is necessary to optimize the algorithms of compression coding in ordertoimprovethecodingefficiency,especiallyinrealtimeapplications.First we introduce the basic theories and standards of video coding, and thebackground of home and abroad. Next we simply introduced the structure of encoderand decoder, key technologies and main performance of H.264 standard, especiallydescribed on the main technologyin video coding layer of H.264.Then we analyze theencoder's performance and computational complexity of H.264 in detail, and researchdeeply on the part of prediction mode selection which has a highly computationalcomplexity.Finally, we completed three aspects of the work as follows in focus: Firstly, weanalyze the fast intra-prediction method in detail. Secondly, we present a fast modeselection method based on threshold by utilizing the relationships of prediction modesof neighbor macroblocks and the coding characters of the relevant prediction modes.Based on the threshold, this algorithm can determine whether the most probable modecan be selected as the final forecasting result, and it excludes massive uselessintra-prediction modes and avoids many unnecessary computations about many costs.Finally, by having many times of coding accomplishment and experiments, andcomparing with the original algorithm, it can be demonstrated that this algorithm candecreases the complexity of the intra-prediction decision effectively, at the same time,it maintains the coding performance fundamentally, and has a favorable adaptiveability.