Underwater vehicles have been widely used in the regular inspection of water-conveyance tunnels. However, for the long-distance water-conveyance tunnels, the high-precision navigation of underwater vehicles is very challenging as there is no satellite navigation signal. In this paper, we propose a tunnel joint extraction method based on deep learning, which can be used for the vision-assisted navigation of underwater vehicles in these tunnels. A light-weight semantic segmentation network for joint extraction is designed with a new architecture which overcomes the class imbalance problem and enables the real-time execution on a moderated GPU. Experimental results demonstrate that our method is effective and can significantly improve the navigation precision in long-distance water-conveyance tunnels.
Joint Extraction Method for Underwater Vehicle Navigation in Long-Distance Water-Conveyance Tunnels
Lect. Notes Electrical Eng.
International Conference on Autonomous Unmanned Systems ; 2021 ; Changsha, China September 24, 2021 - September 26, 2021
Proceedings of 2021 International Conference on Autonomous Unmanned Systems (ICAUS 2021) ; Kapitel : 58 ; 577-584
18.03.2022
8 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Joint Extraction Method for Underwater Vehicle Navigation in Long-Distance Water-Conveyance Tunnels
British Library Conference Proceedings | 2022
|Long-distance water conveyance tunnel underwater inspection robot and method
Europäisches Patentamt | 2020
|TUNNELS - Long distance record set for ROV underwater tunnel inspection
Online Contents | 1999
|