International Display Workshops General Incorporated Association

[3DSAp2-15L] Point Cloud Compression with Optimization on Point Density for Reconstruction of Visually Significant Structure

*Hideaki Kimata1 (1. Kogakuin University (Japan))

point cloud compression, deep learning, compression noise

https://doi.org/10.36463/idw.2023.1404

We propose a compression method of a point cloud, which represents the shape of objects in the real world, while reducing compression noise that severely impairs object features, especially hole-popping. We introduce measures of point density and shape significance for optimizing coding rate and noise.