AquaGS: Fast Underwater Scene Reconstruction with SfM-Free Gaussian Splatting
This addresses the problem of slow and failure-prone underwater reconstruction for robotic platforms, though it appears incremental as it builds on existing techniques like SeaThru, MVS, NeRF, and 3DGS.
The paper tackles underwater scene reconstruction by introducing AquaGS, an SfM-free model that uses SeaThru, MVS, NeRF, and 3D Gaussian Splatting to separate scene details from medium features, achieving high-precision reconstruction in 30 seconds with only 3 image inputs.
Underwater scene reconstruction is a critical tech-nology for underwater operations, enabling the generation of 3D models from images captured by underwater platforms. However, the quality of underwater images is often degraded due to medium interference, which limits the effectiveness of Structure-from-Motion (SfM) pose estimation, leading to subsequent reconstruction failures. Additionally, SfM methods typically operate at slower speeds, further hindering their applicability in real-time scenarios. In this paper, we introduce AquaGS, an SfM-free underwater scene reconstruction model based on the SeaThru algorithm, which facilitates rapid and accurate separation of scene details and medium features. Our approach initializes Gaussians by integrating state-of-the-art multi-view stereo (MVS) technology, employs implicit Neural Radiance Fields (NeRF) for rendering translucent media and utilizes the latest explicit 3D Gaussian Splatting (3DGS) technique to render object surfaces, which effectively addresses the limitations of traditional methods and accurately simulates underwater optical phenomena. Experimental results on the data set and the robot platform show that our model can complete high-precision reconstruction in 30 seconds with only 3 image inputs, significantly enhancing the practical application of the algorithm in robotic platforms.