UAV-Assisted Scan-to-Simulation for Landslides Using Physics-Informed Gaussian Splatting

2026-05-11

AI Summary

This study addresses the limitations of existing landslide simulation methods, which rely on digital elevation models and grid-based representations that lack visual realism and are ill-suited for interactive applications and public education. The authors propose the first end-to-end framework integrating 3D Gaussian Splatting (3DGS) with the Material Point Method (MPM), enabling high-fidelity landslide simulation directly from real-world drone imagery. By introducing physics-informed 3DGS modeling and a low-anisotropy voxel filling strategy, the method achieves seamless reconstruction of real scenes and physically accurate landslide dynamics. Validation on an actual landslide site in Hong Kong demonstrates that the approach simultaneously delivers high visual fidelity and precise physical behavior, significantly outperforming current state-of-the-art techniques.

Abstract

Landslide monitoring and simulation play an important role in urban safety assessment and disaster prevention. Existing landslide simulation pipelines typically rely on digital elevation model and mesh-based representations, which are suitable for geometric analysis, but often lack visual realism. This limitation reduces their effectiveness in interactive applications, hazard communication, and public education. In this paper, we propose a UAV-based scan-to-simulation framework that bridges photorealistic scene capture and physics-based landslide simulation through 3DGS. Specifically, our pipeline includes four stages: (1) UAV-based acquisition of slope imagery, (2) reconstruction of a low-anisotropy 3DGS scene representation, (3) volumetric conversion of the target simulation region by filling the interior of the surface-based model, and (4) integration with the Material Point Method (MPM) for landslide simulation. We validate the proposed framework on a real landslide site in Hong Kong that experienced a severe landslide event. The results show that our method supports both realistic visual reconstruction and effective simulation.

Problem

• landslide simulation
• visual realism
• digital elevation model
• mesh-based representation
• hazard communication

Innovation

• Physics-Informed Gaussian Splatting
• UAV-based 3D reconstruction
• Material Point Method
• Landslide simulation
• Scan-to-simulation