High-Speed Coding Unit Depth Identification Based on Texture Image Information Using SVM.

Abstract

The High-Efficiency Video Coding (HEVC/H.265) standard was created by a group called the Joint Collaborative Team on Video Coding (JCT-VC) and was released in 2013.In HEVC Rate-Distortion Optimization (RDO) is employed for partitioning the coding tree units into coding units recursively due to this the computational complexity increases. However, changing from macroblocks to coding tree units the compression efficiency increases and increases the encoding of the video sequence. This paper presents a Support Vector Machine (SVM)-based method for finding the fastest coding unit division in intra-prediction HEVC without compromising compression efficiency. All partitions of CTU are assessed using five characteristics: Root Mean Square Error (RMSE), Sub CU Complexity Difference (SCCD), Standard Deviation (SD), Directional Complexity (DC), and Quantization Parameter (QP) to optimize the intra-prediction of HEVC in all intra-configurations. The simulation results reveal combining directional complexity and standard deviation yields a more accurate classification. SVM was used to separate split-unsplit data, and the conventional rate-distortion optimisation technique was applied to separate samples that were difficult to separate. The results show that the encoding process has been completed 67.44% faster, albeit with a marginal increase in bit rate.
Subjects
STANDARD deviations; VIDEO coding; MATHEMATICAL optimization; BIT rate; SUPPORT vector machines; IMAGE retrieval

Description