Visual Navigation for Recovering an AUV by Another AUV in Shallow Water

Abstract

Autonomous underwater vehicles (AUVs) play very important roles in underwater missions. However, the reliability of the automated recovery of AUVs has still not been well addressed. We propose a vision-based framework for automatically recovering an AUV by another AUV in shallow water. The proposed framework contains a detection phase for the robust detection of underwater landmarks mounted on the docking station in shallow water and a pose-estimation phase for estimating the pose between AUVs and underwater landmarks. We propose a Laplacian-of-Gaussian-based coarse-to-fine blockwise (LCB) method for the detection of underwater landmarks to overcome ambient light and nonuniform spreading, which are the two main problems in shallow water. We propose a novel method for pose estimation in practical cases where landmarks are broken or covered by biofouling. In the experiments, we show that our proposed LCB method outperforms the state-of-the-art method in terms of remote landmark detection. We then combine our proposed vision-based framework with acoustic sensors in field experiments to demonstrate its effectiveness in the automated recovery of AUVs.

Autonomous Underwater Vehicles: Localization, Navigation, and Communication for Collaborative Missions

“…An experiment using a ship model has been conducted in a laboratory to evaluate the feasibility of the algorithm. The test results demonstrated that the average localization error is approximately 5 cm and the average relative location error is approximately 2% in the range of 3.6 m. A similar approach was proposed by Liu et al [71]. A vision-based framework for automatically recovering an AUV by another AUV in shallow water was presented in this work.…”

Estimation of Positions and Poses of Autonomous Underwater Vehicle Relative to Docking Station Based on Adaptive Extraction of Visual Guidance Features

“…The guidance feature extraction was carried out using the method proposed in this study, and the PPARD estimation results are presented in Table 2. 22) in shallow-water field experiments using the lamps and docking station deployed in [14]. The SIA-3 AUV had a diameter of 384 mm, a length of 5486 mm, and a weight of 1500 kg in air.…”

“…Image segmentation is mainly used to segment guidance lamp regions to extract the guidance features. The processing methods, here, include the mean-shift algorithm [11], region growing method [12], and deep learning [13,14]. Guidance feature extraction can be divided into two types, namely guidance feature extraction based on image binarization (GFEBIB) and guided feature extraction based on edge detection (GFEBED).…”

“…The dataset D recovery [14] was used to study the guidance feature extraction. To solve the above problems, a new method for PPARD estimation based on guidance image preprocessing and the adaptive extraction of visual guidance features was proposed in this study.…”

Neuro-Fuzzy Dynamic Position Prediction for Autonomous Work-Class ROV Docking

“…However, a TMS suspended from a surface platform such as a surface vessel, presents a highly dynamic system, with wave height and period dictating the viability of launch and recovery operations [16].Although docking of UUVs to a moving docking station is reported, the research is mainly focused on an AUV docking and on compensation of disturbances in the horizontal plane (e.g., cross-current), while assuming minimal docking station heave oscillations. Recovering of an AUV by another AUV in shallow water is presented in [17]. The system consists of a "mother" AUV with a funnel shaped docking station attached to its body, designed to accommodate launch and recovery of the "daughter" AUV.…”