基于时空相关性的宏块编码模式预测算法

Macroblock Coding Mode Predictive Method Based on Spatio-Temporal Correlation

与以往的各种视频编码标准相比,H.264视频编码标准在编码效率上获得了最大50%的提高.但这是以提高编码的计算复杂度为代价的,其中运动估计和模式选择的计算复杂度最大.文中提出一种基于时空相关性的宏块编码模式预测算法,该算法首先检查根据时空相关性预测得到的模式,再根据相应的阈值判断是否结束模式选择过程,而不用对所有的模式进行全搜索.试验结果证明:在保持相近率失真性能的同时,该算法可以有效地降低编码的计算复杂度.

H.264 video coding standard can achieve gains in coding efficiency of up to 50% than previous standards. However, this comes at a cost in considerably increased complexity at the encoder mainly due to motion estimation and mode decision. In this paper, a macroblock coding mode predictive method based on spatio-temporal correlation was proposed. This method first searches the modes predicted by the spatio-temporal correlation, then decides whether to terminate the mode decision procedure according to a corresponding threshold instead of searching all modes. Experimental results show that this method can efficiently reduces the computational complexity with similar RD performance.