节点文献
基于H.264/AVC视频编码运动估计算法的优化
Otimization on Motion Estimation Algorithm Based on H.264/AVC
【作者】 白世军;
【导师】 王煜;
【作者基本信息】 哈尔滨工业大学 , 控制科学与工程, 2008, 硕士
【摘要】 H.264/AVC是由ITU-T视频编码专家组VCEG(Video Coding Experts Group)和ISO/IEC运动图象专家组MPEG(Moving Picture Experts Group)共同制定的视频编码标准,这一编码标准可获得很高的编码效率,尤其是在低码率方面比MPEG-4有明显提高,适合低宽带、高质量网络视频应用的需要。为了方便H.264在低码率高实时性应用系统中的实现,须对其编码算法进行优化。经过对H.264编码器的分析可以得出,运动估计是视频压缩编码中的核心技术之一,采用运动估计和运动补偿技术可消除视频信号的时间冗余以提高编码效率。因此,如何提高运动估计的效率,使运动估计算法搜索过程更健壮、更快速、更高效成为目前研究的热点之一。本文首先论述了H.264视频编码标准的基本原理和关键技术;然后介绍了几种已有的经典块匹配运动估计算法,并分析了它们的优缺点;随后本文深入的分析了H.264推荐的运动估计核心算法UMHexagonS,针对该算法运动估计实时性不足的缺点,对该算法进行了四方面的优化,分别是初始搜索点的确定(即参考搜索窗口的位置);参考搜索窗口大小的动态生成;大六边形搜索和小六边形(小钻石)搜索模式。优化后的UMHexagonS算法能够高效快速的定位参考搜索窗口的位置和大小,在整个搜索过程中改进的大小六边形搜索模式可以显著减少搜索点数量并能达到较好的匹配效果。在H.264参考软件模型JM10.1下,经过VC6.0编译,实现该算法,通过对各种测试序列的实验证明,优化后的算法与UMHexagonS算法相比,在重建图象质量和码率接近的情况下,运动估计时间平均节省了18.102%,降低了算法的复杂度,提高了编码器的实时性。
【Abstract】 H.264/AVC is the video coding standard jointly developed by ITU-T Video Coding Experts Group (VCEG) and ISO / IEC Moving Picture Experts Group (MPEG). This coding standard has high coding efficiency which is significantly increased in the low bit-rate compared with MPEG-4, and is available for the low-bandwidth, high-quality network video applications. In order to facilitate the realization of H.264 in the low bit rate and real-time application system the coding algorithm has to be optimized. After the analysis of H.264 encoder we can get that, motion estimation is one of the mainly technology for the video compression coding, and with the motion compensation technology they can eliminate time redundant of video signal to improve the coding efficiency. As a result, how to improve the efficiency of motion estimation to make the search process more robust, faster and more efficient becomes one of the hot spots of the current study.This paper first discusses the basic principles and key technologies of H.264 video coding standard; then introduces several existing classical Block-Matching Motion Estimation and analyses their strengths and weaknesses respectively; whereafter, in depth analyses the H.264 recommended core algorithm of motion estimation– UMHexagonS. Aimed at improving the lack of motion estimation algorithm, four aspects of the algorithm were optimized, which are identification in the initial search point (the location of the reference search range), the dynamically generation of reference search range size as well as big hexagon and small hexagon search mode. Optimized UMHexagonS algorithm can fast and efficient identify the location and size of the reference search range. The improving big and small hexagonal search mode in the entire search process can significantly reduce the number of search point and achieve a better match results.The optimized algorithm is realized in the H.264 reference software model JM10.1, compiled by VC6.0. Experimenting with some typical video sequences proves that, compared to UMHexagonS algorithm, our optimized algorithm can save about 18.102% ME time and reduce the complexity of original scheme as well as enhance the real-time performance of the encoder but has almost no changing in the reconstructed picture quality and bitrate.
【Key words】 H.264/AVC; motion estimation; motion vector; video compression;