A METHOD AND APPARATUS FOR ENCODING/DECODING THE GEOMETRY OF A POINT CLOUD REPRESENTING A 3D OBJECT

摘要

The present principles relates to a two-steps approach for encoding and decoding the geometry of a point cloud. In a first step, an octree-based structure is obtained by splitting recursively a cube encompassing the point cloud until the leaf cubes associated with the leaf nodes of said octree-based structure reach down an expected size. In a second step, for each leaf cube associated with the leaf nodes of said octree-based structure (IO), the approach determines if a local octree-based structure is associated (or not) with a leaf cube by using a Rate-Distortion Optimisation process that optimizes a trade-off between a bit-rate for encoding a candidate octree-based structure approximating the geometry of points of the point cloud which are included in said leaf cube of the octree-based structure, and a distortion that takes into account spatial distances between, on one hand, said points of the point cloud, and on the other hand, points included in leaf cubes associated with leaf nodes of the candidate octree-based structure.