摘要
针对已有可分级视频压缩方法编码复杂度较高的问题,提出一种基于增强残差编码的分布式可分级HEVC压缩方法。在编解码端采用相同的SI(原帧的估算)建立方法来获得最高的压缩率,并使用若干的SI帧来估算EL与SI残差相关性,使用相关性模型决定最低有效位的比特数量并生成校正子,解码端据此校正子重构增强层信息;设计了自适应的编码模式选择机制选择率失真性能较高的编码模式。对比实验结果显示,该方法具有较低的编码复杂度,同时具有较好的率失真性能。
Aimed at the problem of high coding complexity of scalable video compression method, this paper proposed a en- hancement layer residual encoding based distributed scalable HEVC compression method. Firstly, it utilized the same SI crea- ting method by both of eneoder and decoder to obtain the highest compression rate, and used several SI frames to estimate the correlation between EL residual and SI residual. Then, it used the correlation model to decide the count of the least significant bit（LSB） and generated the syndrome, the decoder reconstructed the enhancement layer information based on this syndrome. Lastly, it designed adaptive coding mode selection mechanism to select the coding mode with better RD-rate performance. Comparison experiment results show that the proposal encoding complexity and RD performance outperform the relevant methods.
出处
《计算机应用研究》
CSCD
北大核心
2016年第11期3503-3508,共6页
Application Research of Computers
基金
国家自然科学基金资助项目(61502343)
广西自然科学基金资助项目(2015GXNSFAA139295
2013GXNSFBA019275
2013GXNSFBA019276)
梧州市科技开发资助项目(2015A03006)
广西高校科研资助项目(2013YB227
2013YB228)
武汉科技大学青年科研骨干基金资助项目(2014XZ017)
关键词
可分级视频编码
质量可分级
时间可分级
编码复杂度
相关性模型
scalable video encoding
quality sealability
temporal scalability
encoding complexity
correlation model
