Image Texture Based Fast CU Size Selection Algorithm for HEVC Intra Coding

High Efficiency Video Coding (HEVC) is the newest standard of video coding being developed by JCT-VC. It flexibly uses quad-tree structure, which makes the encoding procedure more complex and time-consuming compared with H.264/AVC. For intra prediction, determining the optimal distribution of coding unit (CU)size requires huge computing time. In this paper, we propose a new fast CU size selection algorithm based on image texture for HEVC intra coding. Some CU sizes can be adaptively skipped due to the characteristic of input image texture. Moreover, the proposed algorithm innovatively adopts the grey level co-occurrence matrix (GLCM) to measure image texture to select CU size. Experimental result shows that this algorithm performs about 26.08% time saving with negligible loss of encoding efficiency.