PolarGS: Polarimetric Cues for Ambiguity-Free Gaussian Splatting with Accurate Geometry Recovery
🤖 AI Summary
3D Gaussian Splatting (3DGS) suffers from geometric inaccuracies in reflective and textureless regions due to photometric ambiguity. To address this, we propose Polarization-Guided 3DGS—the first method to incorporate Degree of Linear Polarization (DoLP) and Angle of Linear Polarization (AoLP) into the 3DGS optimization pipeline. Specifically, DoLP is leveraged to identify reflective regions and guide color refinement, while AoLP/DoLP jointly enhance PatchMatch-based depth completion and Gaussian point cloud densification. Our approach extends the RGB-D 3DGS framework by integrating polarization imaging, polarization decomposition, photometric correction, depth completion, and back-projection fusion. Evaluated across diverse challenging scenes—including highly reflective and low-texture environments—our method achieves显著 improvements in geometric accuracy and completeness. It demonstrates superior robustness and fine-detail recovery in problematic regions, while maintaining full compatibility with mainstream 3DGS architectures.
Application Category
📝 Abstract
Recent advances in surface reconstruction for 3D Gaussian Splatting (3DGS) have enabled remarkable geometric accuracy. However, their performance degrades in photometrically ambiguous regions such as reflective and textureless surfaces, where unreliable cues disrupt photometric consistency and hinder accurate geometry estimation. Reflected light is often partially polarized in a manner that reveals surface orientation, making polarization an optic complement to photometric cues in resolving such ambiguities. Therefore, we propose PolarGS, an optics-aware extension of RGB-based 3DGS that leverages polarization as an optical prior to resolve photometric ambiguities and enhance reconstruction accuracy. Specifically, we introduce two complementary modules: a polarization-guided photometric correction strategy, which ensures photometric consistency by identifying reflective regions via the Degree of Linear Polarization (DoLP) and refining reflective Gaussians with Color Refinement Maps; and a polarization-enhanced Gaussian densification mechanism for textureless area geometry recovery, which integrates both Angle and Degree of Linear Polarization (A/DoLP) into a PatchMatch-based depth completion process. This enables the back-projection and fusion of new Gaussians, leading to more complete reconstruction. PolarGS is framework-agnostic and achieves superior geometric accuracy compared to state-of-the-art methods.
Problem
Research questions and friction points this paper is trying to address.
Resolves photometric ambiguities in reflective surfaces
Enhances geometry recovery in textureless regions
Integrates polarization cues for accurate 3D reconstruction
Innovation
Methods, ideas, or system contributions that make the work stand out.
Polarization-guided photometric correction for reflective surfaces
Polarization-enhanced Gaussian densification for textureless areas
Framework-agnostic integration of polarization cues into 3DGS
🔎 Similar Papers
No similar papers found.
