Efficient Group-Based Packing Strategy for 6DoF Immersive Video Streaming

MPEG immersive video (MIV) standard technology is designed to support immersive video transmissions by removing the redundancy between multiview videos. Inter-view redundancy removal is achieved by packing the input videos into atlases, which are compact representations of multiview videos. Although this process can significantly reduce the pixel rates, it makes immersive video streaming difficult. The MPEG-I community has proposed group-based encoding as a way to improve the objective rendering performance. However, an efficient packing strategy for groups has yet to be investigated. This paper describes an efficient group-based packing strategy that enables immersive video streaming and parallel processing. The goal of this study was to evaluate the effectiveness of group-based encoding by comparing different packing strategies with MIV-encoding parameters. The experiment results show that the proposed packing method achieves better rendering results with a reasonable pixel rate.