preprints.org > engineering > marine engineering > doi: 10.20944/preprints202310.1759.v1

Preprint Article Version 1 Preserved in Portico This version is not peer-reviewed

Version 1 : Received: 26 October 2023 / Approved: 26 October 2023 / Online: 27 October 2023 (12:05:39 CEST)

The port environments have some structures that still need regular control. However, the nature of these environments presents a number of challenges in determining an accurate vehicle position and consequently performing successful image similarity detection. These scenarios, however, can have poor visibility conditions and are highly dynamic, making place recognition difficult because the appearance of a place can change over time. Under these operational conditions, imaging sonars are a promising solution. There are some factors that affect the quality of the captured images, but the fact that they do not suffer from haze is a plus. Therefore, in this paper, a purely acoustic similarity detection method based on forward-looking sonar (FLS) data is proposed to overcome the inherent perceptual problems in a port scenario. Considering the lack of available data in this context and to facilitate the variation of environmental parameters and sensor configurations, a port scenario was simulated using the Stonefish simulator. Experiments were conducted on predefined trajectories near structures. The place recognition approach performs better compared to the results obtained with optical imagery. In general, it provides a good compromise in terms of distinctiveness and achieves a performance of 87.5% when appropriate constraints and assumptions are made.

appearance-based navigation; autonomous underwater vehicles; binary features; BoW; FLS; inspection; loop closure; stonefish

Engineering, Marine Engineering

* All users must log in before leaving a comment