Key Technologies Based On The H.264/avc Video Codec

H.264/AVC video coding standard is a new international standard developed on Mar.2003 by joint video team. Compared with previous video coding standards, such as MPEG-1, MPEG-2, H.263, the new video coding standard showed a better performance in the low transmission rate. It adopts many advanced coding technologies, such as 7 blocks for motion estimation, sub-pixel motion vectors, multiple reference frames. Since the strong advantageous coding techniques, it upgrades almost 50% in the coding efficiency. Yet, all these technologies require a better ability of complexity and computation, which restrict the real time application of H.264/AVC in communication fields. The excellent coding efficiency of H.264/AVC has a great attraction to the market,and the fast coding algorithm and its implementation are the precondition and guarantee for the market application of H.264/AVC, so the optimization of H.264/AVC coding techniques has a great theoretical significance and application values.In this dissertation, firstly, it states the encoding principle of H.264/AVC, together with its main function modules and newly adopted technologies. Secondly, the motion estimation search algorithms are discussed, and then many kinds of classical search algorithms are analysed in comparison. A new fast motion estimation algorithm with prediction center and thresholds for early termination is proposed. Experiments on the proposed method show the encoding speed is largely improved, while the subjective image quality only changes a little. Meanwhile, the proposed method results in a successfully decreasing of the complexity of the encoder and a highly improving of the search efficiency. Thirdly, this dissertation introduces deblocking filter technology in video codec, especially analyses the principle and algorithm of the loop-filter used in H.264/AVC. Through a large number of simulation sequences it is proved that the introduced deblocking filter technology not only can save a lot of bit rates, but also improve subjective image quality. Fourthly, the adaptive loop-filter which has high complexity makes the computational process very consuming.The calculation of Bs (Boundary strength) occupies about 90% of the filtering calculation, therefore this dissertation introduces how to decrease the complexity of Bs to alleviate the time-consuming characteristic of the decoder. Fifthly, although loop-filter technology can effectively remove blocking artifacts, the encoder and decoder must be fully matched, so it lacks flexibility. The proposed deblocking filter algorithm only alters the decoder in the filtering without any change in the encoder, so it can be more flexibly applied. Experiments show the PSNR (Peak Signal Noise Ratio) only have a small change with a little improvement of subjective image quality. Sixthly, integer transform and quantization technologies are also studied in this dissertation.