Self-Supervised Large Scale Point Cloud Completion for Archaeological Site Restoration

2025-03-06

AI Summary

Large-scale archaeological point clouds often suffer from severe incompleteness, fragmented surfaces, and highly non-uniform point distributions, rendering existing self-supervised completion methods inadequate for high-fidelity reconstruction. To address this, we propose a self-supervised point cloud completion framework based on Multi-Center Orthogonal Projection (MCOP): 3D point clouds are projected onto five-channel images (RGB + depth + rotation angle) and formulated as an image inpainting task. We introduce MCOP representation and a patch-level structural reconstruction mechanism, jointly optimized with regularization and structural consistency losses—enabling geometric prior learning without requiring complete ground-truth supervision. Evaluated on over 600 real-world archaeological sites in Peru, our method significantly outperforms state-of-the-art self-supervised approaches, enabling high-precision digital restoration and conservation of damaged cultural heritage.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Non-convex Optimization

Application Category

• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Point cloud completion helps restore partial incomplete point clouds suffering occlusions. Current self-supervised methods fail to give high fidelity completion for large objects with missing surfaces and unbalanced distribution of available points. In this paper, we present a novel method for restoring large-scale point clouds with limited and imbalanced ground-truth. Using rough boundary annotations for a region of interest, we project the original point clouds into a multiple-center-of-projection (MCOP) image, where fragments are projected to images of 5 channels (RGB, depth, and rotation). Completion of the original point cloud is reduced to inpainting the missing pixels in the MCOP images. Due to lack of complete structures and an unbalanced distribution of existing parts, we develop a self-supervised scheme which learns to infill the MCOP image with points resembling existing"complete"patches. Special losses are applied to further enhance the regularity and consistency of completed MCOP images, which is mapped back to 3D to form final restoration. Extensive experiments demonstrate the superiority of our method in completing 600+ incomplete and unbalanced archaeological structures in Peru.

Problem

• Restores large-scale point clouds with limited, imbalanced ground-truth data.
• Addresses high-fidelity completion for large objects with missing surfaces.
• Develops self-supervised method for archaeological site restoration using MCOP images.

Innovation

• Uses MCOP images for point cloud restoration
• Self-supervised learning for infilling missing data
• Special losses enhance regularity and consistency