Coral species identification with texture or structure images using a two-level classifier based on Convolutional Neural Networks

Anabel Gómez-Ríos*, S. Tabik, Julián Luengo, A. S.M. Shihavuddin, Francisco Herrera

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

34 Downloads (Pure)

Abstract

Corals are crucial animals as they support a large part of marine life. The automatic classification of corals species based on underwater images is important as it can help experts to track and detect threatened and vulnerable coral species. However, this classification is complicated due to the nature of coral underwater images and the fact that current underwater coral datasets are unrealistic as they contain only texture images, while the images taken by autonomous underwater vehicles show the complete coral structure. The objective of this paper is two-fold. The first is to build a dataset that is representative of the problem of classifying underwater coral images, the StructureRSMAS dataset. The second is to build a classifier capable of resolving the real problem of classifying corals, based either on texture or structure images. We have achieved this by using a two-level classifier composed of three ResNet models. The first level recognizes whether the input image is a texture or a structure image. Then, the second level identifies the coral species. To do this, we have used a known texture dataset, RSMAS, and StructureRSMAS.
Original languageEnglish
Article number104891
JournalKnowledge-Based Systems
Volume184
Number of pages10
ISSN0950-7051
DOIs
Publication statusPublished - 2019

Keywords

  • Coral images classification
  • Structure coral images
  • Deep learning
  • Convolutional neural networks
  • Inception
  • ResNet
  • DenseNet

Cite this

Gómez-Ríos, Anabel ; Tabik, S. ; Luengo, Julián ; Shihavuddin, A. S.M. ; Herrera, Francisco. / Coral species identification with texture or structure images using a two-level classifier based on Convolutional Neural Networks. In: Knowledge-Based Systems. 2019 ; Vol. 184.
@article{3847ece57cdb4091a15a306c4cbd2416,
title = "Coral species identification with texture or structure images using a two-level classifier based on Convolutional Neural Networks",
abstract = "Corals are crucial animals as they support a large part of marine life. The automatic classification of corals species based on underwater images is important as it can help experts to track and detect threatened and vulnerable coral species. However, this classification is complicated due to the nature of coral underwater images and the fact that current underwater coral datasets are unrealistic as they contain only texture images, while the images taken by autonomous underwater vehicles show the complete coral structure. The objective of this paper is two-fold. The first is to build a dataset that is representative of the problem of classifying underwater coral images, the StructureRSMAS dataset. The second is to build a classifier capable of resolving the real problem of classifying corals, based either on texture or structure images. We have achieved this by using a two-level classifier composed of three ResNet models. The first level recognizes whether the input image is a texture or a structure image. Then, the second level identifies the coral species. To do this, we have used a known texture dataset, RSMAS, and StructureRSMAS.",
keywords = "Coral images classification, Structure coral images, Deep learning, Convolutional neural networks, Inception, ResNet, DenseNet",
author = "Anabel G{\'o}mez-R{\'i}os and S. Tabik and Juli{\'a}n Luengo and Shihavuddin, {A. S.M.} and Francisco Herrera",
year = "2019",
doi = "10.1016/j.knosys.2019.104891",
language = "English",
volume = "184",
journal = "Knowledge-Based Systems",
issn = "0950-7051",
publisher = "Elsevier",

}

Coral species identification with texture or structure images using a two-level classifier based on Convolutional Neural Networks. / Gómez-Ríos, Anabel; Tabik, S.; Luengo, Julián; Shihavuddin, A. S.M.; Herrera, Francisco.

In: Knowledge-Based Systems, Vol. 184, 104891, 2019.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Coral species identification with texture or structure images using a two-level classifier based on Convolutional Neural Networks

AU - Gómez-Ríos, Anabel

AU - Tabik, S.

AU - Luengo, Julián

AU - Shihavuddin, A. S.M.

AU - Herrera, Francisco

PY - 2019

Y1 - 2019

N2 - Corals are crucial animals as they support a large part of marine life. The automatic classification of corals species based on underwater images is important as it can help experts to track and detect threatened and vulnerable coral species. However, this classification is complicated due to the nature of coral underwater images and the fact that current underwater coral datasets are unrealistic as they contain only texture images, while the images taken by autonomous underwater vehicles show the complete coral structure. The objective of this paper is two-fold. The first is to build a dataset that is representative of the problem of classifying underwater coral images, the StructureRSMAS dataset. The second is to build a classifier capable of resolving the real problem of classifying corals, based either on texture or structure images. We have achieved this by using a two-level classifier composed of three ResNet models. The first level recognizes whether the input image is a texture or a structure image. Then, the second level identifies the coral species. To do this, we have used a known texture dataset, RSMAS, and StructureRSMAS.

AB - Corals are crucial animals as they support a large part of marine life. The automatic classification of corals species based on underwater images is important as it can help experts to track and detect threatened and vulnerable coral species. However, this classification is complicated due to the nature of coral underwater images and the fact that current underwater coral datasets are unrealistic as they contain only texture images, while the images taken by autonomous underwater vehicles show the complete coral structure. The objective of this paper is two-fold. The first is to build a dataset that is representative of the problem of classifying underwater coral images, the StructureRSMAS dataset. The second is to build a classifier capable of resolving the real problem of classifying corals, based either on texture or structure images. We have achieved this by using a two-level classifier composed of three ResNet models. The first level recognizes whether the input image is a texture or a structure image. Then, the second level identifies the coral species. To do this, we have used a known texture dataset, RSMAS, and StructureRSMAS.

KW - Coral images classification

KW - Structure coral images

KW - Deep learning

KW - Convolutional neural networks

KW - Inception

KW - ResNet

KW - DenseNet

U2 - 10.1016/j.knosys.2019.104891

DO - 10.1016/j.knosys.2019.104891

M3 - Journal article

VL - 184

JO - Knowledge-Based Systems

JF - Knowledge-Based Systems

SN - 0950-7051

M1 - 104891

ER -