elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

A study on modern deep learning detection algorithms for automatic target recognition in sidescan sonar images

Steiniger, Yannik and Groen, Johannes and Stoppe, Jannis and Kraus, Dieter and Meisen, Tobias (2021) A study on modern deep learning detection algorithms for automatic target recognition in sidescan sonar images. In: 6th Underwater Acoustics Conference and Exhibition, UACE 2021, 44 (1). Acoustical Society of America. 6th Underwater Acoustics Conference & Exhibition, 2021-06-20 - 2021-06-25, online. doi: 10.1121/2.0001470. ISSN 1939-800X.

[img] PDF
4MB

Official URL: https://asa.scitation.org/doi/10.1121/2.0001470

Abstract

State-of-the art deep learning models have shown remarkable performance on computer vision tasks like object classification or detection. These networks are typically trained on large-scale datasets of natural RGB images. However, sidescan sonar images are gray-scaled images representing acoustic intensities. The fundamental differences between camera and sonar as well as the images itself makes it necessary to investigate the transfer of results achieved on RGB images to the sonar imagery domain. Therefore, we compare the deep learning detection algorithm YOLOv2 with its updated version YOLOv3, both adopted for object detection in sidescan sonar images. In addition to this, a small convolutional neural network (CNN) is trained from scratch and used for detection. The experiments answer two questions: First, whether, as for general computer vision problems, transfer learning of large deep learning models is preferable over training of custom networks when dealing with limited sonar data. Secondly, whether improvements in the YOLO architecture, developed based on RGB images, lead to significant improvements on sonar data as well. Our results show that YOLOv3 indeed performs better than YOLOv2. Furthermore, YOLOv3 achieves a true positive rate of up to 98.2% and outperforms the small CNN.

Item URL in elib:https://elib.dlr.de/144633/
Document Type:Conference or Workshop Item (Speech)
Title:A study on modern deep learning detection algorithms for automatic target recognition in sidescan sonar images
Authors:
AuthorsInstitution or Email of AuthorsAuthor's ORCID iDORCID Put Code
Steiniger, YannikUNSPECIFIEDhttps://orcid.org/0000-0002-9327-446XUNSPECIFIED
Groen, JohannesUNSPECIFIEDUNSPECIFIEDUNSPECIFIED
Stoppe, JannisUNSPECIFIEDhttps://orcid.org/0000-0003-2952-3422UNSPECIFIED
Kraus, DieterCity Univertity of Applied Sciences BremenUNSPECIFIEDUNSPECIFIED
Meisen, TobiasUniversity of Wuppertalhttps://orcid.org/0000-0002-1969-559XUNSPECIFIED
Date:15 October 2021
Journal or Publication Title:6th Underwater Acoustics Conference and Exhibition, UACE 2021
Refereed publication:Yes
Open Access:Yes
Gold Open Access:No
In SCOPUS:Yes
In ISI Web of Science:No
Volume:44
DOI:10.1121/2.0001470
Publisher:Acoustical Society of America
ISSN:1939-800X
Status:Published
Keywords:Sidescan sonar, sonar imagery, object detection, automatic target recognition, deep learning
Event Title:6th Underwater Acoustics Conference & Exhibition
Event Location:online
Event Type:international Conference
Event Start Date:20 June 2021
Event End Date:25 June 2021
HGF - Research field:other
HGF - Program:other
HGF - Program Themes:other
DLR - Research area:no assignment
DLR - Program:no assignment
DLR - Research theme (Project):no assignment
Location: Bremerhaven
Institutes and Institutions:Institute for the Protection of Maritime Infrastructures > Maritime Security Technologies
Deposited By: Steiniger, Yannik
Deposited On:25 Oct 2021 12:02
Last Modified:24 Apr 2024 20:43

Repository Staff Only: item control page

Browse
Search
Help & Contact
Information
electronic library is running on EPrints 3.3.12
Website and database design: Copyright © German Aerospace Center (DLR). All rights reserved.