摘要: AVS是中国自主制定的音视频编码标准,在视频编码算法中采用子像素搜索方法,使搜索更精确、压缩效果更好,但算法复杂度较高。基于搜索点失真随着与全局最小点之间距离的增加而增大,以及运动矢量具有中心偏置的特性,该文提出一种三步快速子像素搜索算法。实验结果表明,该算法与标准算法相比,在保证视频质量基本不变的同时,复杂度降低了65%以上。
关键词:
音视频编码标准,
视频编码,
运动估计,
子像素搜索
Abstract: AVS is a domestic audio and video coding standard of China. In its video coding, sub-pixels search method is applied to achieve more accurate motion description and higher compression efficiency, with higher complexity. In order to reduce the complexity, a fast three step sub-pixels search algorithm is presented based on monotonous increase of the distortion of a search point with increasing distance from that search point to the Global Minimum Point(GMP), and the center-biased of motion vector. Simulation results show that the new algorithm can reduce computational complexity by more than 65% with little quality degradation.
Key words:
Audio Video Coding Standard(AVS),
video coding,
motion estimation,
sub-pixels search
中图分类号:
陈永华;王 毅;刘东华;周经野. 用于AVS视频编码的三步快速子像素搜索算法[J]. 计算机工程, 2008, 34(9): 230-231.
CHEN Yong-hua; WANG Yi; LIU Dong-hua; ZHOU Jing-ye. Fast Three Step Sub-pixels Search Algorithm for AVS Video Coding[J]. Computer Engineering, 2008, 34(9): 230-231.