Improved deep learning-based classification of mine-like contacts in sonar images from autonomous underwater vehicles

Abdesselam Bouzerdoum, Philip B. Chapple, Mark Dras, Yi Guo , Len Hamey, Tahereh Hassanzadeh, Thanh Hoang Le, Omid Mohamad Nezami, Mehmet Orgun, Son Lam Phung, Christian Ritz*, Maryam Shahpasand

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contribution

Abstract

This paper describes recent work conducted by a team of researchers from three universities in partnership with defence researchers to investigate deep learning methods for automatic detection of mine-like objects from sidescan sonar images captured by autonomous underwater vehicles. While deep learning can produce state-of-the-art classification performances in several application domains, it often relies on a large amount of labelled training data, which is difficult to obtain in our application. To address this problem, we investigate the use of data augmentation, transfer learning, and compact neural networks. For data augmentation, approaches for increasing the size of the training data are investigated, including standard image processing and manual segmentation. For transfer learning, we use publicly available convolutional neural networks (CNNs) pre-trained on large image datasets, and replace later layers with classifiers trained on sonar image data. For compact neural networks, we train a custom small-sized CNN and also process only the region-of-interest in a sonar snapshot. The proposed techniques are evaluated on a dataset consisting of three classes: mine-like objects, non mine-like objects, and false alarm objects. The experimental results indicate the feasibility of the proposed techniques, with a classification accuracy of 98.3%.
Original languageEnglish
Title of host publication5th Underwater Acoustics Conference and Exhibition 2019, Conference Proceedings
EditorsJohn S. Papadakis
Place of PublicationHeraklion
PublisherInstitute of Applied and Computational Mathematics (IACM)
Pages179-186
Number of pages8
Publication statusPublished - 2019
EventUnderwater Acoustics Conference and Exhibition (5th : 2019) - Hersonissos, Greece
Duration: 30 Jun 20195 Jul 2019

Publication series

Name
ISSN (Electronic)2408-0195

Conference

ConferenceUnderwater Acoustics Conference and Exhibition (5th : 2019)
Abbreviated titleUACE 2019
CountryGreece
CityHersonissos
Period30/06/195/07/19

Keywords

  • automatic target recognition
  • sonar image processing
  • mine-like object detection
  • convolutional neural network

Fingerprint Dive into the research topics of 'Improved deep learning-based classification of mine-like contacts in sonar images from autonomous underwater vehicles'. Together they form a unique fingerprint.

Cite this