Hybrid color attribute compression for point cloud data

Abstract

This paper proposes a color attribute compression method for MPEG Point Cloud Compression (PCC) by exploiting the spatial redundancy among the adjacent points. With the increased interest in representing real-world surface as 3D point clouds, compressing the attributes (i.e., colors and normal directions) of point cloud has attracted great attention in MPEG. The proposed method is based on grouping the adjacent points in blocks. And two encoding modes are supported for each block, which include the run-length encoding mode and palette mode. The final encoding mode for each block is determined through comparing two distortion values based on two encoding modes. Experimental results show that the proposed approach achieves about 28 percent compression ratio than that of MPEG PCC.