Real-Time Visual SLAM for Autonomous Underwater Hull Inspection Using Visual Saliency

TLDR

The paper presents a real‑time monocular visual SLAM algorithm for autonomous underwater hull inspection. The algorithm uses a SLAM navigation prior, selective image registration, and an online bag‑of‑words saliency measure to choose key frames, generate link hypotheses, and detect novelty, thereby addressing limited field‑of‑view and feature‑poor underwater imagery. Experiments on three real‑world hull inspections, including a 3.4‑h/2.7‑km trajectory, demonstrate the effectiveness of the approach.

Abstract

This paper reports a real-time monocular visual simultaneous localization and mapping (SLAM) algorithm and results for its application in the area of autonomous underwater ship hull inspection. The proposed algorithm overcomes some of the specific challenges associated with underwater visual SLAM, namely, limited field of view imagery and feature-poor regions. It does so by exploiting our SLAM navigation prior within the image registration pipeline and by being selective about which imagery is considered informative in terms of our visual SLAM map. A novel online bag-of-words measure for intra and interimage saliency are introduced and are shown to be useful for image key-frame selection, information-gain-based link hypothesis, and novelty detection. Results from three real-world hull inspection experiments evaluate the overall approach, including one survey comprising a 3.4-h/2.7-km-long trajectory.