论文部分内容阅读
针对视频中的视觉感知冗余,提出一种面向视频编码的基于结构分析的感知信号处理模型——结构显著性最小可觉察失真模型.先分策略获取视频帧的有序/无序结构分量和纹理分量进行最小可觉察失真(JND)建模;使用无序结构分量的结构不确定性调节视频帧的显著性区域和JND的对比度掩蔽效应;最后,基于该模型的感知滤波模块被嵌入HEVC中.实验结果表明,在主观视觉感知没有明显损失的情况下,所提出方法的平均码率比HEVC的编码软件HM16.0减少了10.3%.
Aimed at the visual perceived redundancy in video, this paper proposes a structure-aware perceptual signal processing model-structure-aware minimum perceptible distortion model for video coding.Firstly, the strategy is to obtain the ordered / unordered structure components of the video frame and The texture component is modeled as minimum perceptible distortion (JND); the structural uncertainty of the disordered structure component is used to adjust the saliency region of the video frame and the contrast masking effect of JND; finally, the perceptual filtering module based on this model is embedded in HEVC The experimental results show that the average bit rate of the proposed method is reduced by 10.3% compared with that of HEVC encoding software HM16.0 under the condition of no significant loss of subjective visual perception.