.png){kind=icon}
.png){kind=icon}
Return
Multi-objective optimization based perceptual bit allocation for gaming video coding in VVC
.png){kind=icon}
PRE
AI
DOI：10.1016/j.sigpro.2022.108593.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In this paper, a novel multi-objective optimization based coding tree unit (CTU) level perceptual bit allocation method, named MOPBA, for gaming video coding is proposed. To promote the visual quality of gaming video using Versatile Video Coding (VVC), the quality of region of interest (ROI), overall perceptual quality, and quality fluctuation of ROI are optimized simultaneously. More specifically, distortions of ROI and non-ROI (NROI) are weighted based on the foveal characteristics of the human visual system. Bits allocated to ROI and NROI are solved by multi-objective particle swarm optimization (MOPSO). More accurate CTU-level perceptual bit allocation is further resolved by the Newton-Raphson method to increase coding efficiency. Experimental results show that the proposed MOPBA is able to improve the quality of ROI and the perceptual quality of the whole frame while ensures the stability of ROI quality. (C) 2022 Elsevier B.V. All rights reserved.
Keywords:
Perceptual bit allocation
Gaming video coding
Versatile video coding
Multi-objective optimization