Point cloud enhancement by projection sampling

Abstract Next-generation networks are designed to support novel Extended Reality services, which often use point clouds (PCs) to represent real 3D objects. Enhancing the visual quality of PCs by correcting acquisition or transmission errors can significantly improve such services. Herein, we introdu...

Full description

Saved in:
Bibliographic Details
Main Authors: Paolo Giannitrapani, Tiziana Cattai, Stefania Colonnese
Format: Article
Language:English
Published: SpringerOpen 2025-03-01
Series:EURASIP Journal on Image and Video Processing
Subjects:
Online Access:https://doi.org/10.1186/s13640-025-00665-4
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Abstract Next-generation networks are designed to support novel Extended Reality services, which often use point clouds (PCs) to represent real 3D objects. Enhancing the visual quality of PCs by correcting acquisition or transmission errors can significantly improve such services. Herein, we introduce a novel PC enhancement method acting on the PC bidimensional projections computed within the MPEG Video Point Cloud Coding algorithm. We first show that these PC projections are statistically similar to natural images and then we apply to them advanced training-free diffusion sampling. The enhanced projections are then mapped back into the 3D domain, resulting in PC enhancement, which we refer to as Projection Sampling based Point Cloud Enhancement (PS-PCE). We validate PS-PCE on publicly available PCs under various distortions and we compare it with state-of-the-art alternatives. Our results show that the proposed method effectively enhances the PCs, opening new possibilities for PC enhancement in extended reality services.
ISSN:1687-5281