一种质量可伸缩编码中增强层帧间快速算法  

Fast inter prediction algorithm in enhancement layer of quality scalable video coding

作　　者：黄源源[1] 汪大勇[1] 李建平[1] 

机构地区：[1]电子科技大学计算机科学与工程学院,成都611731

出　　处：《电子测量与仪器学报》2011年第2期153-158,共6页Journal of Electronic Measurement and Instrumentation

摘　　要：可伸缩视频编码技术通过增加编码的计算复杂度从而得到较高的编码效率,因此利用可伸缩编码方式自身的特点来提高编码速度成为了当前研究的热点。基于质量可伸缩编码特征,综合考虑基本层和增强层宏块的模式分布、残差系数和运动向量,提出了一种适合于增强层的帧间快速编码算法。实验结果表明,新方法能有效提高编码速度同时不降低编码质量,尤其适用于运动较为复杂的视频序列。Scalable Video Coding has high coding efficiency; however, this enhancement is achieved by increasing computational complexity. Thus it is important to increase the speed of the encoder based on its characteristics. In this paper, a fast inter-prediction algorithm, which is suitable for the coding of enhancement layer, is proposed considering the mode-distribution correlation between the base layer and enhancement layer, residuals, and motion vectors. The experimental results show that the new method greatly improves the speed with negligible loss of coding quality and is especially suitable for the complex video sequences.

关 键 词：可伸缩视频编码 模式分布 残差系数 运动向量 

分 类 号：TN957.52[电子电信—信号与信息处理]