WaterSplat-SLAM: Photorealistic Monocular SLAM in Underwater Environment

TL;DR

WaterSplat-SLAM is a novel underwater monocular SLAM system that provides robust pose estimation and photorealistic dense mapping by integrating semantic filtering and adaptive map management.

Contribution

It introduces a new approach combining semantic medium filtering and an online Gaussian map for high-fidelity underwater environment reconstruction.

Findings

Achieves robust camera tracking in underwater environments.

Produces photorealistic dense maps with high fidelity.

Demonstrates effectiveness on multiple underwater datasets.

Abstract

Underwater monocular SLAM is a challenging problem with applications from autonomous underwater vehicles to marine archaeology. However, existing underwater SLAM methods struggle to produce maps with high-fidelity rendering. In this paper, we propose WaterSplat-SLAM, a novel monocular underwater SLAM system that achieves robust pose estimation and photorealistic dense mapping. Specifically, we couple semantic medium filtering into two-view 3D reconstruction prior to enable underwater-adapted camera tracking and depth estimation. Furthermore, we present a semantic-guided rendering and adaptive map management strategy with an online medium-aware Gaussian map, modeling underwater environment in a photorealistic and compact manner. Experiments on multiple underwater datasets demonstrate that WaterSplat-SLAM achieves robust camera tracking and high-fidelity rendering in underwater environments.