Rapid decision-making method for SCC interframe coding unit mode

Abstract

The invention discloses a rapid decision-making method for an SCC interframe coding unit mode. The rapid decision-making method utilizes information of a Hash interframe search mode of an SCC coding unit and the spatial consistency of coding units, performs statistical analysis on coding unit mode distribution at different depths, and adopts different prediction methods to reduce a prediction mode range so as to ensure the accuracy rate of prediction. At different depths, results of interframe search based on Hash and information of the adjacent CU are utilized to reduce a coding unit prediction mode candidate range, thereby avoiding unnecessary mode searching process. For the coding unit with the depth of 0, whether to skip Inter2Nx2N, Skip, interframe segmentation prediction mode and an intraframe prediction mode is predicted; and for the coding unit with the depth of 1, whether to skip Inter2Nx2N, an interframe segmentation prediction mode and an intraframe prediction mode is predicted. The rapid decision-making method can effectively reduce the mode searching range of the coding units, so as to reduce the complexity of SCC interframe coding.

Description

technical field [0001] The present invention relates to the field of screen content coding (Screen Content Coding, SCC), in particular to the decision-making technology of coding unit mode in SCC inter-frame coding. Background technique [0002] As video conferencing, remote desktop control and other applications become more and more widespread, people's demand for screen videos such as animations and images with text and charts is increasing. The current SCC standard based on the HEVC (High Efficiency Video Coding) video compression coding standard Still being perfected, SCC continues to use the quadtree structure coding unit division method adopted by the HEVC coding standard. The size of the coding unit (CodingUnit, CU) is 64x64, 32x32, 16x16 and 8x8, corresponding to the depth level of the CU is 0 , 1, 2, 3. Among them, the CU with a depth of 0 is recorded as LCU (LargestCodingUnit), and the CU with a depth of 3 is recorded as SCU (SmallestCodingUnit). Considering that...

AI Technical Summary

Problems solved by technology

[0003] Hash-based inter-frame search can effectively improve the efficiency of SCC encoders. For details, see Document 2 (LiB, XuJ, WuF. Aunified framework of hash-based matching for screen content coding [C] / / Visual Communications and Image Processing Conference, 2014IEEE.IEEE, 2014:530-533.), but other The introduction of the encoding mode increases the complexity of the encoder, so the SCC encoder still has a high complexity and needs to be further optimized

Images