Vessel Bearing Estimation Using Visible and Thermal Imaging

Loading...
Thumbnail Image

Access rights

openAccess

URL

Journal Title

Journal ISSN

Volume Title

A4 Artikkeli konferenssijulkaisussa

Date

2023

Major/Subject

Mcode

Degree programme

Language

en

Pages

9
373-381

Series

Image Analysis - 23rd Scandinavian Conference, SCIA 2023, Proceedings, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Volume 13886 LNCS

Abstract

Maritime awareness and autonomous navigation can be enabled by state-of-the-art deep learning methods, by monitoring and tracking the position of maritime vessels. In our experiment, we acquire ship dataset from Megastar cruise ferry campaigns in Baltic Sea. We detect the nearby vessels using visible and infrared imaging sensors and object detectors You-Only-Look-Once v5 (YOLOv5), and Detectron2 RCNN network and use that information along with DeepSORT method to track the position of the vessel. We obtain the bearing of the vessels detected from both infrared and visible image sequences of ship and fuse them using a Kalman filter (KF) and Rauch-Tung-Striebel (RTS) smoother. We then compare the result to the bearing obtained from the automatic identification system (AIS) of the vessel, and also compare it among object detectors. We obtained a root-mean-square error of 0.17∘ in vessel bearing tracking as compared to AIS.

Description

Funding Information: We thank European Space Agency (ESA) for funding and Tallink Megastar crew for enabling the measurements onboard the vessel. We also thank Henrik Ramm-Schmidt from Fleetrange, Toni Hammarberg, and Martta-Kaisa Olkkonen from Finnish Geospatial Research Institute (FGI) for aiding in collecting the data from the campaign. Publisher Copyright: © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Keywords

Bearing estimation, Maritime awareness, Ship detection

Other note

Citation

Gorad, A, Hassan, S S & Särkkä, S 2023, Vessel Bearing Estimation Using Visible and Thermal Imaging . in R Gade, M Felsberg & J-K Kämäräinen (eds), Image Analysis - 23rd Scandinavian Conference, SCIA 2023, Proceedings . Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13886 LNCS, Springer, pp. 373-381, Scandinavian Conference on Image Analysis, Kittilä, Finland, 18/04/2023 . https://doi.org/10.1007/978-3-031-31438-4_25