基于区域划分的深度视频快速编码算法

多视点彩色视频与深度视频作为多媒体系统中主流3D场景表示方式，吸引了越来越多人的关注。深度视频反映场景的几何信息，如何对其进行快速编码尤为重要。提出了一种基于区域划分的深度视频快速编码算法。首先，根据深度视频的边缘和运动属性把深度视频分为四个区域;然后，深入分析不同区域内宏块模式分布比例以及参考帧选择特性，设计了不同的宏块模式选择和参考帧搜索策略来提高编码速度;最后，测试提出算法的编码时间、码率以及合成虚拟视点的质量。实验结果表明，提出算法在保证虚拟视点图像质量和编码码率基本不变的情况下，节约了85.73%～91.06%的编码时间。

Fast depth video coding algorithm based on region division

As the main scheme of 3D scene representation, multiview video plus depth attracts more and more attention. Depth video reflects the geometric information of the scene. It is important to design fast depth video encoding algorithm. A fast depth video coding algorithm based on region division was proposed. Firstly, the depth video was divided into four regions according to the features of edge and motion. Then, macroblock distribution proportion and multi-reference frame selection feature of different regions were analyzed. Consequently, different macroblock mode decision and reference frame selection methods were utilized to speedup depth video encoding. Finally, some experiments were conducted to estimate the proposed algorithm in terms of encoding time, bit rate and virtual view quality. Experimental results show that the proposed algorithm saves encoding time ranging from 85.73% to 91.06% while it maintains virtual view quality and bit rate.