ShadowGS: Shadow-Aware 3D Gaussian Splatting for Satellite Imagery

TL;DR

ShadowGS introduces a physics-based, shadow-aware 3D Gaussian Splatting framework that improves 3D reconstruction accuracy and shadow modeling in satellite imagery, especially under varying illumination and sparse views.

Contribution

It presents a novel shadow-aware 3D Gaussian Splatting method incorporating a physics-based rendering model, shadow consistency constraints, and a shadow map prior for enhanced satellite image reconstruction.

Findings

Outperforms state-of-the-art in shadow decoupling and 3D reconstruction

Achieves high-quality novel view synthesis with minimal training time

Demonstrates robustness across various satellite imaging modalities

Abstract

3D Gaussian Splatting (3DGS) has emerged as a novel paradigm for 3D reconstruction from satellite imagery. However, in multi-temporal satellite images, prevalent shadows exhibit significant inconsistencies due to varying illumination conditions. To address this, we propose ShadowGS, a novel framework based on 3DGS. It leverages a physics-based rendering equation from remote sensing, combined with an efficient ray marching technique, to precisely model geometrically consistent shadows while maintaining efficient rendering. Additionally, it effectively disentangles different illumination components and apparent attributes in the scene. Furthermore, we introduce a shadow consistency constraint that significantly enhances the geometric accuracy of 3D reconstruction. We also incorporate a novel shadow map prior to improve performance with sparse-view inputs. Extensive experiments demonstrate that ShadowGS outperforms current state-of-the-art methods in shadow decoupling accuracy, 3D reconstruction precision, and novel view synthesis quality, with only a few minutes of training. ShadowGS exhibits robust performance across various settings, including RGB, pansharpened, and sparse-view satellite inputs.