Deep Image Representations for Coral Image Classification

Ammar Mahmood, Mohammed Bennamoun, Senjian An, Ferdous A. Sohel, Farid Boussaid, Renae Hovey, Gary A. Kendrick, Robert B. Fisher

Research output: Contribution to journalArticle

1 Citation (Scopus)
30 Downloads (Pure)

Abstract

Healthy coral reefs play a vital role in maintaining biodiversity in tropical marine ecosystems. Remote imaging techniques have facilitated the scientific investigations of these intricate ecosystems, particularly at depths beyond 10 m where SCUBA diving techniques are not time or cost efficient. With millions of digital images of the seafloor collected using remotely operated vehicles and autonomous underwater vehicles (AUVs), manual annotation of these data by marine experts is a tedious, repetitive, and time-consuming task. It takes 10–30 min for a marine expert to meticulously annotate a single image. Automated technology to monitor the health of the oceans would allow for transformational ecological outcomes by standardizing methods to detect and identify species. This paper aims to automate the analysis of large available AUV imagery by developing advanced deep learning tools for rapid and large-scale automatic annotation of marine coral species. Such an automated technology would greatly benefit marine ecological studies in terms of cost, speed, and accuracy. To this end, we propose a deep learning based classification method for coral reefs and report the application of the proposed technique to the automatic annotation of unlabeled mosaics of the coral reef in the Abrolhos Islands, W.A., Australia. Our proposed method automatically quantified the coral coverage in this region and detected a decreasing trend in coral population, which is in line with conclusions drawn by marine ecologists.

Original languageEnglish
Number of pages11
JournalIEEE Journal of Oceanic Engineering
DOIs
Publication statusE-pub ahead of print - 12 Jan 2018

Fingerprint

Reefs
Image classification
Autonomous underwater vehicles
Remotely operated vehicles
Aquatic ecosystems
Biodiversity
Ecosystems
Costs
Health
Imaging techniques
Deep learning

Cite this

@article{39eee6ffd6dd4a8889db54ab287eab4d,
title = "Deep Image Representations for Coral Image Classification",
abstract = "Healthy coral reefs play a vital role in maintaining biodiversity in tropical marine ecosystems. Remote imaging techniques have facilitated the scientific investigations of these intricate ecosystems, particularly at depths beyond 10 m where SCUBA diving techniques are not time or cost efficient. With millions of digital images of the seafloor collected using remotely operated vehicles and autonomous underwater vehicles (AUVs), manual annotation of these data by marine experts is a tedious, repetitive, and time-consuming task. It takes 10–30 min for a marine expert to meticulously annotate a single image. Automated technology to monitor the health of the oceans would allow for transformational ecological outcomes by standardizing methods to detect and identify species. This paper aims to automate the analysis of large available AUV imagery by developing advanced deep learning tools for rapid and large-scale automatic annotation of marine coral species. Such an automated technology would greatly benefit marine ecological studies in terms of cost, speed, and accuracy. To this end, we propose a deep learning based classification method for coral reefs and report the application of the proposed technique to the automatic annotation of unlabeled mosaics of the coral reef in the Abrolhos Islands, W.A., Australia. Our proposed method automatically quantified the coral coverage in this region and detected a decreasing trend in coral population, which is in line with conclusions drawn by marine ecologists.",
keywords = "Australia, Classification, coral population, corals, deep learning, Ecosystems, Feature extraction, Image color analysis, marine ecosystems, marine images, Sociology, Statistics",
author = "Ammar Mahmood and Mohammed Bennamoun and Senjian An and Sohel, {Ferdous A.} and Farid Boussaid and Renae Hovey and Kendrick, {Gary A.} and Fisher, {Robert B.}",
year = "2018",
month = "1",
day = "12",
doi = "10.1109/JOE.2017.2786878",
language = "English",
journal = "IEEE Journal of Oceanic Engineering",
issn = "0364-9059",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",

}

TY - JOUR

T1 - Deep Image Representations for Coral Image Classification

AU - Mahmood, Ammar

AU - Bennamoun, Mohammed

AU - An, Senjian

AU - Sohel, Ferdous A.

AU - Boussaid, Farid

AU - Hovey, Renae

AU - Kendrick, Gary A.

AU - Fisher, Robert B.

PY - 2018/1/12

Y1 - 2018/1/12

N2 - Healthy coral reefs play a vital role in maintaining biodiversity in tropical marine ecosystems. Remote imaging techniques have facilitated the scientific investigations of these intricate ecosystems, particularly at depths beyond 10 m where SCUBA diving techniques are not time or cost efficient. With millions of digital images of the seafloor collected using remotely operated vehicles and autonomous underwater vehicles (AUVs), manual annotation of these data by marine experts is a tedious, repetitive, and time-consuming task. It takes 10–30 min for a marine expert to meticulously annotate a single image. Automated technology to monitor the health of the oceans would allow for transformational ecological outcomes by standardizing methods to detect and identify species. This paper aims to automate the analysis of large available AUV imagery by developing advanced deep learning tools for rapid and large-scale automatic annotation of marine coral species. Such an automated technology would greatly benefit marine ecological studies in terms of cost, speed, and accuracy. To this end, we propose a deep learning based classification method for coral reefs and report the application of the proposed technique to the automatic annotation of unlabeled mosaics of the coral reef in the Abrolhos Islands, W.A., Australia. Our proposed method automatically quantified the coral coverage in this region and detected a decreasing trend in coral population, which is in line with conclusions drawn by marine ecologists.

AB - Healthy coral reefs play a vital role in maintaining biodiversity in tropical marine ecosystems. Remote imaging techniques have facilitated the scientific investigations of these intricate ecosystems, particularly at depths beyond 10 m where SCUBA diving techniques are not time or cost efficient. With millions of digital images of the seafloor collected using remotely operated vehicles and autonomous underwater vehicles (AUVs), manual annotation of these data by marine experts is a tedious, repetitive, and time-consuming task. It takes 10–30 min for a marine expert to meticulously annotate a single image. Automated technology to monitor the health of the oceans would allow for transformational ecological outcomes by standardizing methods to detect and identify species. This paper aims to automate the analysis of large available AUV imagery by developing advanced deep learning tools for rapid and large-scale automatic annotation of marine coral species. Such an automated technology would greatly benefit marine ecological studies in terms of cost, speed, and accuracy. To this end, we propose a deep learning based classification method for coral reefs and report the application of the proposed technique to the automatic annotation of unlabeled mosaics of the coral reef in the Abrolhos Islands, W.A., Australia. Our proposed method automatically quantified the coral coverage in this region and detected a decreasing trend in coral population, which is in line with conclusions drawn by marine ecologists.

KW - Australia

KW - Classification

KW - coral population

KW - corals

KW - deep learning

KW - Ecosystems

KW - Feature extraction

KW - Image color analysis

KW - marine ecosystems

KW - marine images

KW - Sociology

KW - Statistics

UR - http://www.scopus.com/inward/record.url?scp=85041184436&partnerID=8YFLogxK

U2 - 10.1109/JOE.2017.2786878

DO - 10.1109/JOE.2017.2786878

M3 - Article

JO - IEEE Journal of Oceanic Engineering

JF - IEEE Journal of Oceanic Engineering

SN - 0364-9059

ER -