AQUA-SLAM: Tightly-Coupled Underwater Acoustic-Visual-Inertial SLAM with Sensor Calibration
Shida Xu, Kaicheng Zhang, Sen Wang
AI summary
Problem
Visual SLAM systems fail in underwater environments due to poor visibility, marine snow, and frequent feature loss, while existing multi-sensor fusion approaches lack rigorous Doppler velocity log modeling and reliable online calibration, causing significant localization drift.
Approach
The authors develop a tightly coupled graph optimization framework that fuses Doppler velocity log, stereo camera, and IMU measurements, integrated with a fast linear approximation algorithm for real-time extrinsic and transducer calibration.
Key results
- First tightly coupled graph-based SLAM system integrating DVL, camera, and IMU for underwater use
- Novel online calibration algorithm for multisensor extrinsics and DVL transducer misalignment
- Successful validation in tank experiments and North Sea offshore trials
- Superior localization accuracy and robustness compared to state-of-the-art underwater and visual-inertial SLAM systems
Why it matters
Enables reliable autonomous navigation and mapping for underwater robots in challenging offshore environments, advancing marine robotics and infrastructure inspection.
Abstract
Underwater environments pose significant challenges for visual simultaneous localization and mapping (SLAM) systems due to limited visibility, inadequate illumination, and sporadic loss of structural features in images. Addressing these challenges, this article introduces a novel, tightly coupled acoustic-visual- inertial SLAM approach, termed AQUA-SLAM, to fuse a Doppler velocity log (DVL), a stereo camera, and an inertial measurement unit (IMU) within a graph optimization framework. Moreover, we propose an efficient sensor calibration technique, encompassing the multisensor extrinsic calibration (among the DVL, camera, and IMU)andtheDVLtransducermisalignmentcalibration,withafast linear approximation procedure for real-time online execution. The proposed methods are extensively evaluated in a tank environment with ground truth, and validated for offshore applications in the North Sea. The results demonstrate that our method surpasses cur- rent state-of-the-art underwater and visual-inertial SLAM systems in terms of localization accuracy and robustness. The proposed system will be made open-source for the community.