基于3D-HEVC标准的相邻块视差矢量获取算法质量优化的研究

发布时间：2018-02-22 13:51

本文关键词： 高效视频编码标准的三维扩展视点间纹理图编码相邻块视差矢量获取比例组合分组搜索　出处：《电子学报》2017年08期 　论文类型：期刊论文

【摘要】：在高效视频编码标准的三维扩展方案(3D-HEVC)中,相邻块视差矢量获取算法是视点间纹理图编码的一项关键技术.针对现有快速算法获取的视差矢量质量有所下降的问题,本文提出一种利用分组搜索和按比例组合策略的相邻块视差矢量获取质量优化算法.首先,改变原有相邻块视差矢量获取算法中将第一个搜索到的视差矢量作为最终视差矢量的准则;其次,设计一种新的分组结合方法,将时间方向和空间方向上的待搜索的候选块分为若干小组,并将互相临近的候选块中得到的视差矢量按比例结合为新的视差矢量;最后,设计新的分组搜索策略,在新的视差矢量与原有矢量中获取最终相邻块视差矢量.实验结果表明,在维持现有基于编码单元的简化相邻块视差矢量获取算法效率的前提下可实现编码质量提升.
[Abstract]:In 3D-HEVC3D-HEVC, an efficient 3D video coding standard, the parallax vector acquisition algorithm for adjacent blocks is a key technique in interview texture image coding. In this paper, a quality optimization algorithm based on block parallax vector based on grouping search and proportional combination strategy is proposed. The original parallax vector acquisition algorithm of adjacent blocks takes the first parallax vector as the criterion of the final parallax vector. Secondly, a new grouping combination method is designed. The candidate blocks in the direction of time and space are divided into several groups, and the parallax vectors obtained from adjacent candidate blocks are proportionally combined into new parallax vectors. Finally, a new grouping search strategy is designed. Finally, the parallax vector of adjacent blocks is obtained from the new parallax vector and the original vector. The experimental results show that the coding quality can be improved on the premise of maintaining the efficiency of the existing simplified parallax vector acquisition algorithm based on coding unit.
【作者单位】：北京工业大学电子信息与控制工程学院下一代信息网络北京实验室;
【基金】：国家关键技术研发计划(No.2011BAC12B03) 北京市自然科学基金暨教委重点科技项目(No.KZ201610005007) 国家自然科学基金面上项目(No.61672064)
【分类号】：TN919.81

【相似文献】