Multiview Geometric Regularization of Gaussian Splatting for Accurate Radiance Fields

TL;DR

This paper introduces a multiview geometric regularization method for Gaussian Splatting that combines MVS depth, RGB, and normal constraints to improve the accuracy and reliability of 3D scene reconstructions.

Contribution

It proposes a novel multiview regularization strategy integrating MVS-derived depth and normal constraints into Gaussian Splatting, enhancing geometric accuracy and rendering quality.

Findings

Improved geometric accuracy in diverse scenes.

Enhanced rendering quality with regularization.

Effective integration of MVS depth into Gaussian Splatting.

Abstract

Recent methods, such as 2D Gaussian Splatting and Gaussian Opacity Fields, have aimed to address the geometric inaccuracies of 3D Gaussian Splatting while retaining its superior rendering quality. However, these approaches still struggle to reconstruct smooth and reliable geometry, particularly in scenes with significant color variation across viewpoints, due to their per-point appearance modeling and single-view optimization constraints. In this paper, we propose an effective multiview geometric regularization strategy that integrates multiview stereo (MVS) depth, RGB, and normal constraints into Gaussian Splatting initialization and optimization. Our key insight is the complementary relationship between MVS-derived depth points and Gaussian Splatting-optimized positions: MVS robustly estimates geometry in regions of high color variation through local patch-based matching and epipolar constraints, whereas Gaussian Splatting provides more reliable and less noisy depth estimates near object boundaries and regions with lower color variation. To leverage this insight, we introduce a median depth-based multiview relative depth loss with uncertainty estimation, effectively integrating MVS depth information into Gaussian Splatting optimization. We also propose an MVS-guided Gaussian Splatting initialization to avoid Gaussians falling into suboptimal positions. Extensive experiments validate that our approach successfully combines these strengths, enhancing both geometric accuracy and rendering quality across diverse indoor and outdoor scenes.