An evaluation of supervised and unsupervised classification techniques for marine benthic habitat mapping using multibeam echosounder data

Jay Calvert, James Asa Strong, Matthew Service, C McGonigle, R Quinn

Research output: Contribution to journalArticle

21 Citations (Scopus)

Abstract

Marine habitat mapping provides information on seabed substrata and faunal community structure to users including research scientists, conservation organizations, and policy makers. Full-coverage acoustic data are frequently used for habitat mapping in combination with video groundtruth data in either a supervised or unsupervised classification. In this investigation, video ground-truth data with a camera footprint of 1 m2 were classified to level 4 of the European Nature Information System habitat classification scheme. Acoustic data with a horizontal resolution of 1 m2 were collected over an area of 130 km2 using a multibeam echosounder, and processed to provide bathymetry and backscatter data. Bathymetric derivatives including eastness, northness, slope, topographic roughness index, vector rugosity measure, and two measures of curvature were created. A feature selection process based on Kruskal–Wallis and post hoc pairwise testing was used to select environmental variables able to discriminate ground-truth classes. Subsequently, three datasets were formed: backscatter alone (BS), backscatter combined with bathymetry and derivatives (BSDER), and bathymetry and derivatives alone (DER). Two classifications were performed on each of the datasets to produce habitat maps: maximum likelihood supervised classification (MLC) and ISO Cluster unsupervised classification. Accuracy of the supervised habitat maps was assessed using total agreement, quantity disagreement, and allocation disagreement. Agreement in the unsupervised maps was assessed using the Cramer’s V coefficient. Choice of input data produced large differences in the accuracy of the supervised maps, but did not have the same effect on the unsupervised maps. Accuracies were 46, 56, and 49% when calculated using the sample and 52, 65, and 51% when using an unbiased estimate of the population for the BS, BSDER, and DER maps, respectively. Cramer’s V was 0.371, 0.417, and 0.366 for the BS, BSDER, and DER maps, respectively.
LanguageEnglish
Pages1498-1513
JournalICES Journal of Marine Science
Volume72
DOIs
Publication statusPublished - 12 Nov 2015

Fingerprint

unsupervised classification
image classification
backscatter
habitat
habitats
bathymetry
acoustic data
acoustics
topographic slope
methodology
information systems
roughness
cameras
community structure
evaluation
footprint
curvature
environmental factors
information system
testing

Keywords

  • habitat mapping
  • multibeam echosounder
  • supervised classification
  • towed video
  • unsupervised classification

Cite this

@article{5fc5e1d0a60d429e8dbf377cf6421562,
title = "An evaluation of supervised and unsupervised classification techniques for marine benthic habitat mapping using multibeam echosounder data",
abstract = "Marine habitat mapping provides information on seabed substrata and faunal community structure to users including research scientists, conservation organizations, and policy makers. Full-coverage acoustic data are frequently used for habitat mapping in combination with video groundtruth data in either a supervised or unsupervised classification. In this investigation, video ground-truth data with a camera footprint of 1 m2 were classified to level 4 of the European Nature Information System habitat classification scheme. Acoustic data with a horizontal resolution of 1 m2 were collected over an area of 130 km2 using a multibeam echosounder, and processed to provide bathymetry and backscatter data. Bathymetric derivatives including eastness, northness, slope, topographic roughness index, vector rugosity measure, and two measures of curvature were created. A feature selection process based on Kruskal–Wallis and post hoc pairwise testing was used to select environmental variables able to discriminate ground-truth classes. Subsequently, three datasets were formed: backscatter alone (BS), backscatter combined with bathymetry and derivatives (BSDER), and bathymetry and derivatives alone (DER). Two classifications were performed on each of the datasets to produce habitat maps: maximum likelihood supervised classification (MLC) and ISO Cluster unsupervised classification. Accuracy of the supervised habitat maps was assessed using total agreement, quantity disagreement, and allocation disagreement. Agreement in the unsupervised maps was assessed using the Cramer’s V coefficient. Choice of input data produced large differences in the accuracy of the supervised maps, but did not have the same effect on the unsupervised maps. Accuracies were 46, 56, and 49{\%} when calculated using the sample and 52, 65, and 51{\%} when using an unbiased estimate of the population for the BS, BSDER, and DER maps, respectively. Cramer’s V was 0.371, 0.417, and 0.366 for the BS, BSDER, and DER maps, respectively.",
keywords = "habitat mapping, multibeam echosounder, supervised classification, towed video, unsupervised classification",
author = "Jay Calvert and Strong, {James Asa} and Matthew Service and C McGonigle and R Quinn",
year = "2015",
month = "11",
day = "12",
doi = "10.1093/icesjms/fsu223",
language = "English",
volume = "72",
pages = "1498--1513",
journal = "ICES Journal of Marine Science",
issn = "1054-3139",

}

An evaluation of supervised and unsupervised classification techniques for marine benthic habitat mapping using multibeam echosounder data. / Calvert, Jay; Strong, James Asa; Service, Matthew; McGonigle, C; Quinn, R.

In: ICES Journal of Marine Science, Vol. 72, 12.11.2015, p. 1498-1513.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An evaluation of supervised and unsupervised classification techniques for marine benthic habitat mapping using multibeam echosounder data

AU - Calvert, Jay

AU - Strong, James Asa

AU - Service, Matthew

AU - McGonigle, C

AU - Quinn, R

PY - 2015/11/12

Y1 - 2015/11/12

N2 - Marine habitat mapping provides information on seabed substrata and faunal community structure to users including research scientists, conservation organizations, and policy makers. Full-coverage acoustic data are frequently used for habitat mapping in combination with video groundtruth data in either a supervised or unsupervised classification. In this investigation, video ground-truth data with a camera footprint of 1 m2 were classified to level 4 of the European Nature Information System habitat classification scheme. Acoustic data with a horizontal resolution of 1 m2 were collected over an area of 130 km2 using a multibeam echosounder, and processed to provide bathymetry and backscatter data. Bathymetric derivatives including eastness, northness, slope, topographic roughness index, vector rugosity measure, and two measures of curvature were created. A feature selection process based on Kruskal–Wallis and post hoc pairwise testing was used to select environmental variables able to discriminate ground-truth classes. Subsequently, three datasets were formed: backscatter alone (BS), backscatter combined with bathymetry and derivatives (BSDER), and bathymetry and derivatives alone (DER). Two classifications were performed on each of the datasets to produce habitat maps: maximum likelihood supervised classification (MLC) and ISO Cluster unsupervised classification. Accuracy of the supervised habitat maps was assessed using total agreement, quantity disagreement, and allocation disagreement. Agreement in the unsupervised maps was assessed using the Cramer’s V coefficient. Choice of input data produced large differences in the accuracy of the supervised maps, but did not have the same effect on the unsupervised maps. Accuracies were 46, 56, and 49% when calculated using the sample and 52, 65, and 51% when using an unbiased estimate of the population for the BS, BSDER, and DER maps, respectively. Cramer’s V was 0.371, 0.417, and 0.366 for the BS, BSDER, and DER maps, respectively.

AB - Marine habitat mapping provides information on seabed substrata and faunal community structure to users including research scientists, conservation organizations, and policy makers. Full-coverage acoustic data are frequently used for habitat mapping in combination with video groundtruth data in either a supervised or unsupervised classification. In this investigation, video ground-truth data with a camera footprint of 1 m2 were classified to level 4 of the European Nature Information System habitat classification scheme. Acoustic data with a horizontal resolution of 1 m2 were collected over an area of 130 km2 using a multibeam echosounder, and processed to provide bathymetry and backscatter data. Bathymetric derivatives including eastness, northness, slope, topographic roughness index, vector rugosity measure, and two measures of curvature were created. A feature selection process based on Kruskal–Wallis and post hoc pairwise testing was used to select environmental variables able to discriminate ground-truth classes. Subsequently, three datasets were formed: backscatter alone (BS), backscatter combined with bathymetry and derivatives (BSDER), and bathymetry and derivatives alone (DER). Two classifications were performed on each of the datasets to produce habitat maps: maximum likelihood supervised classification (MLC) and ISO Cluster unsupervised classification. Accuracy of the supervised habitat maps was assessed using total agreement, quantity disagreement, and allocation disagreement. Agreement in the unsupervised maps was assessed using the Cramer’s V coefficient. Choice of input data produced large differences in the accuracy of the supervised maps, but did not have the same effect on the unsupervised maps. Accuracies were 46, 56, and 49% when calculated using the sample and 52, 65, and 51% when using an unbiased estimate of the population for the BS, BSDER, and DER maps, respectively. Cramer’s V was 0.371, 0.417, and 0.366 for the BS, BSDER, and DER maps, respectively.

KW - habitat mapping

KW - multibeam echosounder

KW - supervised classification

KW - towed video

KW - unsupervised classification

U2 - 10.1093/icesjms/fsu223

DO - 10.1093/icesjms/fsu223

M3 - Article

VL - 72

SP - 1498

EP - 1513

JO - ICES Journal of Marine Science

T2 - ICES Journal of Marine Science

JF - ICES Journal of Marine Science

SN - 1054-3139

ER -