Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural Network

This paper proposes a sonar-based underwater object classification method for autonomous underwater vehicles (AUVs) by reconstructing an object’s three-dimensional (3D) geometry. The point cloud of underwater objects can be generated from sonar images captured while the AUV passes over the object. T...

Full description

Bibliographic Details
Main Authors: Minsung Sung, Jason Kim, Hyeonwoo Cho, Meungsuk Lee, Son-Cheol Yu
Format: Article
Language:English
Published: MDPI AG 2020-10-01
Series:Electronics
Subjects:
AUV
Online Access:https://www.mdpi.com/2079-9292/9/11/1763
id doaj-b6f8483280154a10a58d2e7909e8b84f
record_format Article
spelling doaj-b6f8483280154a10a58d2e7909e8b84f2020-11-25T03:37:09ZengMDPI AGElectronics2079-92922020-10-0191763176310.3390/electronics9111763Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural NetworkMinsung Sung0Jason Kim1Hyeonwoo Cho2Meungsuk Lee3Son-Cheol Yu4Department of IT Engineering, Pohnag University of Science and Technology, Pohang 37673, KoreaDepartment of IT Engineering, Pohnag University of Science and Technology, Pohang 37673, KoreaDepartment of IT Engineering, Pohnag University of Science and Technology, Pohang 37673, KoreaDepartment of Electrical Engineering, Pohnag University of Science and Technology, Pohang 37673, KoreaDepartment of IT Engineering, Pohnag University of Science and Technology, Pohang 37673, KoreaThis paper proposes a sonar-based underwater object classification method for autonomous underwater vehicles (AUVs) by reconstructing an object’s three-dimensional (3D) geometry. The point cloud of underwater objects can be generated from sonar images captured while the AUV passes over the object. Then, a neural network can predict the class given the generated point cloud. By reconstructing the 3D shape of the object, the proposed method can classify the object accurately through a straightforward training process. We verified the proposed method by performing simulations and field experiments.https://www.mdpi.com/2079-9292/9/11/1763AUVforward-scan sonarsonar classificationsonar sensorunderwater object classification
collection DOAJ
language English
format Article
sources DOAJ
author Minsung Sung
Jason Kim
Hyeonwoo Cho
Meungsuk Lee
Son-Cheol Yu
spellingShingle Minsung Sung
Jason Kim
Hyeonwoo Cho
Meungsuk Lee
Son-Cheol Yu
Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural Network
Electronics
AUV
forward-scan sonar
sonar classification
sonar sensor
underwater object classification
author_facet Minsung Sung
Jason Kim
Hyeonwoo Cho
Meungsuk Lee
Son-Cheol Yu
author_sort Minsung Sung
title Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural Network
title_short Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural Network
title_full Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural Network
title_fullStr Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural Network
title_full_unstemmed Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural Network
title_sort underwater-sonar-image-based 3d point cloud reconstruction for high data utilization and object classification using a neural network
publisher MDPI AG
series Electronics
issn 2079-9292
publishDate 2020-10-01
description This paper proposes a sonar-based underwater object classification method for autonomous underwater vehicles (AUVs) by reconstructing an object’s three-dimensional (3D) geometry. The point cloud of underwater objects can be generated from sonar images captured while the AUV passes over the object. Then, a neural network can predict the class given the generated point cloud. By reconstructing the 3D shape of the object, the proposed method can classify the object accurately through a straightforward training process. We verified the proposed method by performing simulations and field experiments.
topic AUV
forward-scan sonar
sonar classification
sonar sensor
underwater object classification
url https://www.mdpi.com/2079-9292/9/11/1763
work_keys_str_mv AT minsungsung underwatersonarimagebased3dpointcloudreconstructionforhighdatautilizationandobjectclassificationusinganeuralnetwork
AT jasonkim underwatersonarimagebased3dpointcloudreconstructionforhighdatautilizationandobjectclassificationusinganeuralnetwork
AT hyeonwoocho underwatersonarimagebased3dpointcloudreconstructionforhighdatautilizationandobjectclassificationusinganeuralnetwork
AT meungsuklee underwatersonarimagebased3dpointcloudreconstructionforhighdatautilizationandobjectclassificationusinganeuralnetwork
AT soncheolyu underwatersonarimagebased3dpointcloudreconstructionforhighdatautilizationandobjectclassificationusinganeuralnetwork
_version_ 1724546792885846016