对目前三维高效视频编码(3D-HEVC)标准的性能进行了分析,针对其深度图编码采用的合成视点失真变化(SVDC)计算复杂度高的问题,提出了一种基于自适应分段跳过规则的视点合成优化(VSO)方法.考虑到原始方法跳过了深度数据没有变化的模块或像素行,但忽略了可以在VSO中跳过的较小区域,该方法首先根据深度零失真准则和纹理平滑度准则判别出深度图中不会造成合成视点失真的区域,然后在深度图编码进行SVDC计算前基于自适应分段规则来提前跳过上述零失真区域,最后终止该区域的SVDC计算过程,有效地降低编码复杂度.对比实验结果证明了该方法的有效性,能在保持编码质量不变的前提下,平均减少14.2％的编码时间.%The performance of the standard of three dimensional high efficiency video coding ( 3 D-HEVC ) is analyzed , and aiming at its problem that the complexity of calculating synthesized view distortion change ( SVDC ) in depth coding is high , a view synthesized optimization ( VSO ) method based on adaptive segmentation skipping is pro-posed .Considering that the original method skips pixels of lines or blocks without the change of the depth data , but ignores the smaller areas that can also be skipped in VSO , the proposed method determines the region of the depth map that does not cause the synthesized view distortion according to the depth zero-distortion criterion and the tex-ture smoothing criterion .Then , the zero-distortion region is skipped based on adaptive segmentation skip before the SVDC calculation of the depth map coding .Finally, the SVDC calculation process for the area is terminated .The experimental results show that the proposed method can reduce the coding time by 14.2％while keeping the coding quality unchanged .
展开▼
