Complexity Scalable Algorithm Of Inter Modes Decision In H.264/AVC

With the rapid development of wireless communications and multimedia communication technology, all kinds of video application platform mature gradually in the process of industrialization. Herein, video communication faces a big challenge from matching between high complexity of video coding and different computation capacity of various terminals, so it is necessary to enable scalable complexity of video coding. The generation video coding standard H.264/AVC adopt intra coding, multi-size block segmentation, multiple reference frames and other new technologies, which make its compression property and network adaptability improved greatly, but also make its complexity increased immensely. On that basis, studied inter mode selection algorithm and complexity scalable methods deeply in connection with the complexity of H.264/AVC's multi-size block segmentation.On the basis of the analysis of the process and algorithm's complexity of inter mode selection,and the study of the existied complexity scalability algorithms,this paper proposed a novel algorithm-TSC(Inter Modes Decision of Complexity Scalability Based on Temporal and Spatial Correlation).It utilized the spatial correlation and temporal correlation, adopted complexity scalability factor,hence the complexity can change flexibly from 20% to 100%. The main innovation is mainly reflected in the following:(1)Utilized the spatial correlation and temporal correlation of the adjacent macro block, analysed statistically the order of priority of predicted modes in different combinations of adjacent macroblock. Inter modes predicted in turn of priority scale, make sure the maximum possible mode can be the one which be calculation. even if the number of predicted modes decrease.(2)In the precess of inter modes participate prediction, adopted complexity scalability factor, controlled the number of inter modes involved in, made the computing resources are concentrated in the areas which's spatial correlation and temporal correlation is weakly, and optimized allocation of resource.Experimental results show that the proposed algorithm can achieve the tradeoff between the video quality and its computation complexity, adapt the different computing power of different equipments.