Abstract: The 3D video extension of High Efficiency Video Coding（3D-HEVC） is the latest video coding standard for MVD. In order to reduce the computational complexity of depth map intra coding, an algorithm is proposed to reduce the number of modes in the full-RD search list. By using the self-correlation between the optimal mode and the first mode in the full-RD search list, few modes are chose and added to the full-RD search list. Experimental results show that the proposed algorithm can save 26.35% of depth map encoding time on average with negligible distortion for the synthesized views compared with HTM 13.0.

Key words: 3D High Efficiency Video Coding（3D-HEVC）, depth maps, intra coding, computational complexity, optimal mode, self-correlation

摘要： 3D-HEVC是多视点纹理加深度视频（MVD）的最新编码标准。针对3D-HEVC中深度图帧内编码过程计算复杂度高的问题，提出了一种减少帧内候选模式数目的快速模式决策算法。该算法基于最佳预测模式与粗选模式列表首位模式与之间的相关性，选择成为最佳预测模式概率较高的几种HEVC帧内预测模式加入到粗选模式列表中。实验结果表明，与原始HTM13.0相比，该算法使得深度图的编码时间减少26.35%，而引起的合成视点的失真忽略不计。

关键词: 3D高效视频编码（3D-HEVC）, 深度图, 帧内编码, 计算复杂度, 最佳预测模式, 相关性