Distance Weighted Refining Segmentation Method for Visual Quality Improvement in V-PCC
Min Ku Lee, Yong-Hwan Kim · 2023
In this paper, we propose an innovative method for refining segmentation method that improves the visual quality of Video-based Point Cloud Compression (V-PCC) encoder. Recently standardized as an international standard by MPEG, V-PCC standard provides state-of-the-art performance in compressing dynamic and dense point cloud object. However, lossy V-PCC encoder has an unavoidable problem of visual quality degradation due to lost points. When converting a 3D point cloud to 2D patches in the V-PCC encoder, some points constituting a point cloud are not converted. In particular, in the refining segmentation of the 2D patch generation process, points that are changed the projection plane due to over-smoothing can be discarded. We propose a distance weighted refining segmentation method that reduces the number of missed points to improve visual quality. Experimental results show a noticeable improvement in visual quality with minor coding gain.