基于时空相关与纹理特性的HEVC编码单元快速划分算法

Fast Coding Unit Partition Algorithm for HEVC Based on Temporal-spatial Correlation and Texture Property

为了降低高效视频编码(HEVC)中编码单元(CU)四叉树递归划分的复杂度,提出了一种基于时空相关性和纹理特性的CU快速划分算法。首先,利用视频中存在的时空相关性,根据自适应权值预测当前CU深度范围。然后,将深度范围中满足纹理特性的CU提前终止划分。实验表明,在全I帧模式下论文算法与HM10.0相比,在保证编码性能的同时,帧内编码时间平均缩减52.41%。

In order to reduce the complexity of recursive partition of quadtree-based coding unit(CU)in high efficiency video coding(HEVC),a fast algorithm for CU partition based on temporal-spatial correlation and texture property is proposed.Firstly,the depth range of current CU is predicted based on the adaptive weights using the temporal-spatial correlation existing in the video.Then,texture properties are used to suspend some CU division of corresponding size in advance.On the basis of the experimental results,compared with HM10.0,the proposed algorithm can save 52.41%coding time on average,while maintaining high video coding quality.