Points2Surf: Learning Implicit Surfaces from Point Cloud Patches

TL;DR

Points2Surf tackles the challenge of reconstructing watertight surfaces from noisy, unstructured point clouds without normals by learning an implicit surface representation. It decomposes the signed distance function into a coarse global sign and a detailed local distance, implemented via separate encoders for local patches and global context, enabling robust generalization to shapes with diverse geometry and topology. The method yields consistent qualitative and quantitative gains over both non-data-driven SPR and data-driven baselines on unseen classes, at the cost of longer inference times and some minor topological noise. This approach advances practical surface reconstruction for real-world scanning workflows by improving detail preservation and resilience to noise in diverse object categories.

Abstract

A key step in any scanning-based asset creation workflow is to convert unordered point clouds to a surface. Classical methods (e.g., Poisson reconstruction) start to degrade in the presence of noisy and partial scans. Hence, deep learning based methods have recently been proposed to produce complete surfaces, even from partial scans. However, such data-driven methods struggle to generalize to new shapes with large geometric and topological variations. We present Points2Surf, a novel patch-based learning framework that produces accurate surfaces directly from raw scans without normals. Learning a prior over a combination of detailed local patches and coarse global information improves generalization performance and reconstruction accuracy. Our extensive comparison on both synthetic and real data demonstrates a clear advantage of our method over state-of-the-art alternatives on previously unseen classes (on average, Points2Surf brings down reconstruction error by 30% over SPR and by 270%+ over deep learning based SotA methods) at the cost of longer computation times and a slight increase in small-scale topological noise in some cases. Our source code, pre-trained model, and dataset are available on: https://github.com/ErlerPhilipp/points2surf

Figures (9)

• Figure 1: We present Points2Surf, a method to reconstruct an accurate implicit surface from a noisy point cloud. Unlike current data-driven surface reconstruction methods like DeepSDF and AtlasNet, it is patch-based, improves detail reconstruction, and unlike Screened Poisson Reconstruction (SPR), a learned prior of low-level patch shapes improves reconstruction accuracy. Note the quality of reconstructions, both geometric and topological, against the original surfaces. The ability of Points2Surf to generalize to new shapes makes it the first learning-based approach with significant generalization ability under both geometric and topological variations.
• Figure 2: Points2Surf Architecture. Given a query point $x$ (red) and a point cloud $P$ (gray), we sample a local patch (yellow) and a coarse global subsample (purple) of the point cloud. These are encoded into two feature vectors that are fed to a decoder, which outputs a logit of the sign probability and the absolute distance of the SDF at the query point $x$.
• Figure 3: Dataset examples. Examples of the ABC dataset and its three variants are shown on the left, examples of the famous dataset and its five variants on the right.
• Figure 4: Qualitative comparison of surface reconstructions. We evaluate one example from each dataset variant with each method. Colors show the distance of the reconstructed surface to the ground-truth surface.
• Figure 5: Additional qualitative comparison of surface reconstructions on the ABC dataset. We evaluate two examples from each dataset variant with each method. Colors show the distance of the reconstructed surface to the ground truth surface.