Fast CU Partition and Intra Mode Decision Method for H.266/VVC

Qiuwen Zhang,Lixun Huang,Bin Jiang,Yihan Wang

doi:10.1109/access.2020.3004580

Qiuwen Zhang, Lixun Huang + Show 2 more

Open Access

https://doi.org/10.1109/access.2020.3004580

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2020
Citations: 65	License type: CC BY 4.0

Affiliation: Zhengzhou University of Light Industry

Abstract

The Versatile Video Coding (H.266/VVC) standard has developed by Joint Video Exploration Team (JVET). Compared with the previous generation video coding standard, the H.266/VVC is more outstanding. Since the H.266/VVC introduces multi-type tree (MTT) structure including binary tree (BT) and ternary tree (TT), which brings the significant coding efficiency but increases coding complexity. Moreover, the intra prediction modes have increased from 35 to 67, which can provide more accurate prediction than H.265/High Efficiency Video Coding (HEVC). Therefore, these can improve the encoding quality, but increase computational complexity. To reduce the computational complexity, this paper designs a fast coding unit (CU) partition and intra mode decision algorithm, which includes fast CU partition based on random forest classifier (RFC) model and fast intra prediction modes optimization based on texture region features. Simulation results indicate that the proposed scheme can save 54.91% encoding time with only 0.93% increase in BDBR.

Highlights

With the emergence of video applications, such as 4K/8K ultra high definition (UHD), the amount of video data has exploded and the higher requirements have been placed on encoding technology [1]
To reduce the computational complexity and encoding time, this paper proposes a fast coding unit (CU) partition and intra mode decision algorithm, which includes fast CU partition based on random forest classifier (RFC) model and fast intra prediction modes optimization based on texture region features
The remaining of this paper is organized as follows: In Section 2, the proposed fast CU partition and intra mode decision method is described in detail, which includes fast CU partition method based on RFC model and fast intra prediction modes optimization based on texture region features

Summary

Introduction

With the emergence of video applications, such as 4K/8K ultra high definition (UHD), the amount of video data has exploded and the higher requirements have been placed on encoding technology [1]. A fast intra prediction modes decision scheme based on machine learning is developed in [9], which utilizes RFC model to accelerate the coding speed. A fast intra prediction method based on texture complexity is proposed in [17], which can effectively decrease the computational complexity of H.265/HEVC.

Results

Conclusion