Vessel Bearing Estimation Using Visible and Thermal Imaging

Ajinkya Gorad; Syeda Sakira Hassan; Simo Särkkä

doi:10.1007/978-3-031-31438-4_25

Vessel Bearing Estimation Using Visible and Thermal Imaging

Ajinkya Gorad^*, Syeda Sakira Hassan, Simo Särkkä

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

85 Downloads (Pure)

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.

Original language	English
Title of host publication	Image Analysis - 23rd Scandinavian Conference, SCIA 2023, Proceedings
Editors	Rikke Gade, Michael Felsberg, Joni-Kristian Kämäräinen
Publisher	Springer
Pages	373-381
Number of pages	9
ISBN (Electronic)	978-3-031-31438-4
ISBN (Print)	978-3-031-31437-7
DOIs	https://doi.org/10.1007/978-3-031-31438-4_25
Publication status	Published - 2023
MoE publication type	A4 Conference publication
Event	Scandinavian Conference on Image Analysis - Levi, Kittilä, Finland Duration: 18 Apr 2023 → 21 Apr 2023 Conference number: 23

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13886 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	Scandinavian Conference on Image Analysis
Abbreviated title	SCIA
Country/Territory	Finland
City	Kittilä
Period	18/04/2023 → 21/04/2023

Keywords

Bearing estimation
Maritime awareness
Ship detection

Access to Document

10.1007/978-3-031-31438-4_25

Vessel_bearing_estimation_using_visible_and_thermal_imagingAccepted author manuscript, 3.04 MB

Cite this

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 (pp. 373-381). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13886 LNCS). Springer. https://doi.org/10.1007/978-3-031-31438-4_25

Gorad, Ajinkya ; Hassan, Syeda Sakira ; Särkkä, Simo. / Vessel Bearing Estimation Using Visible and Thermal Imaging. Image Analysis - 23rd Scandinavian Conference, SCIA 2023, Proceedings. editor / Rikke Gade ; Michael Felsberg ; Joni-Kristian Kämäräinen. Springer, 2023. pp. 373-381 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{20af69df38574deb9675fcfa76df3e92,

title = "Vessel Bearing Estimation Using Visible and Thermal Imaging",

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.",

keywords = "Bearing estimation, Maritime awareness, Ship detection",

author = "Ajinkya Gorad and Hassan, {Syeda Sakira} and Simo S{\"a}rkk{\"a}",

note = "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: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; Scandinavian Conference on Image Analysis, SCIA ; Conference date: 18-04-2023 Through 21-04-2023",

year = "2023",

doi = "10.1007/978-3-031-31438-4_25",

language = "English",

isbn = "978-3-031-31437-7",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "373--381",

editor = "Rikke Gade and Michael Felsberg and Joni-Kristian K{\"a}m{\"a}r{\"a}inen",

booktitle = "Image Analysis - 23rd Scandinavian Conference, SCIA 2023, Proceedings",

address = "Germany",

}

Gorad, A, Hassan, SS & 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

Vessel Bearing Estimation Using Visible and Thermal Imaging. / Gorad, Ajinkya; Hassan, Syeda Sakira; Särkkä, Simo.
Image Analysis - 23rd Scandinavian Conference, SCIA 2023, Proceedings. ed. / Rikke Gade; Michael Felsberg; Joni-Kristian Kämäräinen. Springer, 2023. p. 373-381 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13886 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Vessel Bearing Estimation Using Visible and Thermal Imaging

AU - Gorad, Ajinkya

AU - Hassan, Syeda Sakira

AU - Särkkä, Simo

N1 - Conference code: 23

PY - 2023

Y1 - 2023

N2 - 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.

AB - 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.

KW - Bearing estimation

KW - Maritime awareness

KW - Ship detection

UR - http://www.scopus.com/inward/record.url?scp=85161447263&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-31438-4_25

DO - 10.1007/978-3-031-31438-4_25

M3 - Conference article in proceedings

AN - SCOPUS:85161447263

SN - 978-3-031-31437-7

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 373

EP - 381

BT - Image Analysis - 23rd Scandinavian Conference, SCIA 2023, Proceedings

A2 - Gade, Rikke

A2 - Felsberg, Michael

A2 - Kämäräinen, Joni-Kristian

PB - Springer

T2 - Scandinavian Conference on Image Analysis

Y2 - 18 April 2023 through 21 April 2023

ER -

Gorad A, Hassan SS, Särkkä S. Vessel Bearing Estimation Using Visible and Thermal Imaging. In Gade R, Felsberg M, Kämäräinen JK, editors, Image Analysis - 23rd Scandinavian Conference, SCIA 2023, Proceedings. Springer. 2023. p. 373-381. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-31438-4_25

Vessel Bearing Estimation Using Visible and Thermal Imaging

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Science-IT

Cite this

Vessel Bearing Estimation Using Visible and Thermal Imaging

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Equipment

Science-IT

Cite this