Projection-based Occupancy Map Coding for 3D Point Cloud Compression

Abstract

An occupancy map coding method to compress 3D point cloud data is proposed. A 3D point cloud can be coded by projecting into a texture, depth, and occupancy map. This paper presents an efficient way to signal consecutive fully occupied occupancy coding units with a single bit for each coding unit row of an object in the occupancy map is presented. The results showed that the proposed algorithm could save 3.02% of bits without coding loss compared to an existing algorithm.