2D-SuGaR: Surface-Aware Gaussian Splatting for Geometrically Accurate Mesh Reconstruction

TL;DR

This paper improves 2D Gaussian Splatting for surface reconstruction by integrating monocular depth and normal priors, leading to more accurate and robust mesh reconstructions from multi-view images.

Contribution

It introduces a depth-guided initialization and clustering-based pruning for 2DGS, enhancing geometric accuracy and robustness over previous methods.

Findings

Achieves state-of-the-art mesh reconstruction results on DTU dataset.

Maintains high-quality novel view synthesis.

Improves robustness against poor SfM initializations.

Abstract

3D Gaussian Splatting (3DGS) has emerged as a powerful technique for generating photorealistic renderings of a scene in real-time. However, the volumetric nature of 3DGS limits its ability to accurately capture surface geometry. To address this, 2D Gaussian Splatting (2DGS) was proposed to enable view-consistent and geometrically accurate surface reconstruction from multi-view images. However, 2DGS can be sensitive to the initialization of the Gaussian primitives. Reliance on Structure-from-Motion (SfM) initializations, which can produce poor estimates on challenging image sets, may lead to subpar results. In this work, we enhance 2DGS by incorporating monocular depth and normal priors to improve both geometric accuracy and robustness. We propose a depth-guided initialization strategy for Gaussians and introduce a clustering-based technique for pruning degenerate Gaussians. We evaluate our method on the DTU dataset, where it achieves state-of-the-art results in mesh reconstruction while preserving high-quality novel view synthesis.