节点文献

H.264视频编码的帧内预测算法研究

Research and Implementation of H.264 Intra-prediction

【作者】 曹雁

【导师】 刘宁钟;

【作者基本信息】 南京航空航天大学 , 计算机应用技术, 2008, 硕士

【摘要】 H.264是由ITU-T的视频编码专家组和ISO的运动图象专家组共同制定的新的视频编码标准。H.264标准中包含了很多先进的视频压缩编码方法,与以前的视频编码标准相比具有更好的压缩性能,同时由于其设计的合理性,使得该标准具有良好的网络友好性。然而,高编码压缩率是以高的计算复杂度为代价的,这样很难将H.264标准应用到实时性要求较高的场合,限制了标准的应用范围。为了使H.264能够更好地应用于实时多媒体通信,就必须对其算法进行优化以降低计算复杂度,从而缩短编码时间。本文在详细研究H.264视频编码特点的基础上,深入分析了帧内预测编码,并提出了关于帧内预测编码的改进算法。H.264中的帧内预测是充分利用视频信号在空间上的相关性进行压缩编码传输的。标准中采用基于率失真优化的全搜索算法进行最佳预测模式的选择,这样计算量巨大。本文首先分析了基于边缘方向直方图的Pan算法,并进行了优化,实验结果表明,改进的算法在保证图象编码质量和码率的同时,大大缩短了编码时间。其次,深入分析了另外两种基于图象纹理特性的帧内预测算法,并针对其不足之处提出了改进算法。实验表明,该算法在保证图象质量的条件下,编码时间进一步缩短,并且易于进行硬件实现。

【Abstract】 H.264 is the newest video coding standard of the ITU-T Video Coding Experts Group (VCEG) and the ISO/IEC Moving Picture Experts Group (MPEG). H.264 adopted many effective methods to improve compression performance.So it can achieve considerable higher coding efficiency than previous standards.Besides, H.264 processes the provision of“network -friendly”.However, the high complexity comes with the high compression, so it is difficult to implement in real-time applications.Thus, we must optimize the algorithm to reduce the complexity.The article provides an overview of technical features of H.264, introduce the theory of intra prediction in detail, and advance two novel improved fast algorithms of intra prediction.Intra prediction of H.264 utilizes the spatial relativity to compress. H.264 adoptes Full Search algorithm based on Rate Distortion Optimization (RDO) to select the best predictive mode.In this way, the computation is drastically increased. The article,firstly, optimize the Pan’s algorithm, and the simulation results show that without considerable performance degradation, the proposed method decreases complexity of the intra-prediction mode decision.Secondly, the paper analyze another two methods of edge detection, and proposed its improvement algorithm. Experimental results show that the proposed algorithm reduces coding time without considerable performance degradation.Besides, it is easily to apply to hardware.

节点文献中: