3D POINT CLOUD COMPRESSION SYSTEM BASED ON MULTI-SCALE STRUCTURED DICTIONARY LEARNING

摘要

The present invention provides a 3D point cloud compression system based on multi-scale structured dictionary learning. A point cloud data partition module outputs a voxel set obtained by point cloud partitioning and a voxel block set of different scales; a geometric information encoding module outputs an encoded geometric information bit stream; a geometric information decoding module outputs decoded geometric information; an attribute signal encoding module outputs a sparsely encoded coefficient matrix and a learned multi-scale structured dictionary; an attribute signal encoding module outputs the learned multi-scale structured dictionary; an attribute signal compression module outputs a compressed attribute signal bit stream; an attribute signal decoding module outputs a decoded attribute signal; and a 3D point cloud reconstruction module completes reconstruction. The present invention is applicable to lossless geometric and lossy attribute compression of point cloud signals, and by using a natural layer division structure of the point cloud signal, in a direction of signal scales from being coarse to being fine, the reconstruction quality of high-frequency detail information is improved in a gradient manner, thereby being able to obtain a significant performance gain.