Coral classification with hybrid feature representations

© 2016 IEEE. Coral reefs exhibit significant within-class variations, complex between-class boundaries and inconsistent image clarity. This makes coral classification a challenging task. In this paper, we report the application of generic CNN representations combined with hand-crafted features for c...

Full description

Bibliographic Details
Main Authors: Mahmood, A., Bennamoun, M., An, Senjian, Sohel, F., Boussaid, F., Hovey, R., Kendrick, G., Fisher, R.
Format: Conference Paper
Published: 2016
Online Access:http://hdl.handle.net/20.500.11937/70267
_version_ 1848762259386400768
author Mahmood, A.
Bennamoun, M.
An, Senjian
Sohel, F.
Boussaid, F.
Hovey, R.
Kendrick, G.
Fisher, R.
author_facet Mahmood, A.
Bennamoun, M.
An, Senjian
Sohel, F.
Boussaid, F.
Hovey, R.
Kendrick, G.
Fisher, R.
author_sort Mahmood, A.
building Curtin Institutional Repository
collection Online Access
description © 2016 IEEE. Coral reefs exhibit significant within-class variations, complex between-class boundaries and inconsistent image clarity. This makes coral classification a challenging task. In this paper, we report the application of generic CNN representations combined with hand-crafted features for coral reef classification to take advantage of the complementary strengths of these representation types. We extract CNN based features from patches centred at labelled pixels at multiple scales. We use texture and color based hand-crafted features extracted from the same patches to complement the CNN features. Our proposed method achieves a classification accuracy that is higher than the state-of-art methods on the MLC benchmark dataset for corals.
first_indexed 2025-11-14T10:44:43Z
format Conference Paper
id curtin-20.500.11937-70267
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T10:44:43Z
publishDate 2016
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-702672018-08-08T04:56:26Z Coral classification with hybrid feature representations Mahmood, A. Bennamoun, M. An, Senjian Sohel, F. Boussaid, F. Hovey, R. Kendrick, G. Fisher, R. © 2016 IEEE. Coral reefs exhibit significant within-class variations, complex between-class boundaries and inconsistent image clarity. This makes coral classification a challenging task. In this paper, we report the application of generic CNN representations combined with hand-crafted features for coral reef classification to take advantage of the complementary strengths of these representation types. We extract CNN based features from patches centred at labelled pixels at multiple scales. We use texture and color based hand-crafted features extracted from the same patches to complement the CNN features. Our proposed method achieves a classification accuracy that is higher than the state-of-art methods on the MLC benchmark dataset for corals. 2016 Conference Paper http://hdl.handle.net/20.500.11937/70267 10.1109/ICIP.2016.7532411 restricted
spellingShingle Mahmood, A.
Bennamoun, M.
An, Senjian
Sohel, F.
Boussaid, F.
Hovey, R.
Kendrick, G.
Fisher, R.
Coral classification with hybrid feature representations
title Coral classification with hybrid feature representations
title_full Coral classification with hybrid feature representations
title_fullStr Coral classification with hybrid feature representations
title_full_unstemmed Coral classification with hybrid feature representations
title_short Coral classification with hybrid feature representations
title_sort coral classification with hybrid feature representations
url http://hdl.handle.net/20.500.11937/70267