Comparing Solo Versus Ensemble Convolutional Neural Networks for Wetland Classification Using Multi-Spectral Satellite Imagery
Wetlands are important ecosystems that are linked to climate change mitigation. As 25% of global wetlands are located in Canada, accurate and up-to-date wetland classification is of high importance, nationally and internationally. The advent of deep learning techniques has revolutionized the current...
Published in: | Remote Sensing |
---|---|
Main Authors: | , , , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
MDPI AG
2021
|
Subjects: | |
Online Access: | https://doi.org/10.3390/rs13112046 https://doaj.org/article/924bddcdfb254a38885592d6bedf0cae |
id |
ftdoajarticles:oai:doaj.org/article:924bddcdfb254a38885592d6bedf0cae |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:924bddcdfb254a38885592d6bedf0cae 2023-05-15T17:22:43+02:00 Comparing Solo Versus Ensemble Convolutional Neural Networks for Wetland Classification Using Multi-Spectral Satellite Imagery Ali Jamali Masoud Mahdianpari Brian Brisco Jean Granger Fariba Mohammadimanesh Bahram Salehi 2021-05-01T00:00:00Z https://doi.org/10.3390/rs13112046 https://doaj.org/article/924bddcdfb254a38885592d6bedf0cae EN eng MDPI AG https://www.mdpi.com/2072-4292/13/11/2046 https://doaj.org/toc/2072-4292 doi:10.3390/rs13112046 2072-4292 https://doaj.org/article/924bddcdfb254a38885592d6bedf0cae Remote Sensing, Vol 13, Iss 2046, p 2046 (2021) deep learning wetland mapping convolutional neural network satellite image classification ensemble learning Science Q article 2021 ftdoajarticles https://doi.org/10.3390/rs13112046 2022-12-31T16:18:31Z Wetlands are important ecosystems that are linked to climate change mitigation. As 25% of global wetlands are located in Canada, accurate and up-to-date wetland classification is of high importance, nationally and internationally. The advent of deep learning techniques has revolutionized the current use of machine learning algorithms to classify complex environments, specifically in remote sensing. In this paper, we explore the potential and possible limitations to be overcome regarding the use of ensemble deep learning techniques for complex wetland classification and discusses the potential and limitation of various solo convolutional neural networks (CNNs), including DenseNet, GoogLeNet, ShuffleNet, MobileNet, Xception, Inception-ResNet, ResNet18, and ResNet101 in three different study areas located in Newfoundland and Labrador, Canada (i.e., Avalon, Gros Morne, and Grand Falls). Moreover, to improve the classification accuracies of wetland classes of bog, fen, marsh, swamp, and shallow water, the results of the three best CNNs in each study area is fused using three supervised classifiers of random forest (RF), bagged tree (BTree), Bayesian optimized tree (BOT), and one unsupervised majority voting classifier. The results suggest that the ensemble models, in particular BTree, have a valuable role to play in the classification of wetland classes of bog, fen, marsh, swamp, and shallow water. The ensemble CNNs show an improvement of 9.63–19.04% in terms of mean producer’s accuracy compared to the solo CNNs, to recognize wetland classes in three different study areas. This research indicates a promising potential for integrating ensemble-based learning and deep learning for operational large area land cover, particularly complex wetland type classification. Article in Journal/Newspaper Newfoundland Directory of Open Access Journals: DOAJ Articles Newfoundland Canada Remote Sensing 13 11 2046 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
deep learning wetland mapping convolutional neural network satellite image classification ensemble learning Science Q |
spellingShingle |
deep learning wetland mapping convolutional neural network satellite image classification ensemble learning Science Q Ali Jamali Masoud Mahdianpari Brian Brisco Jean Granger Fariba Mohammadimanesh Bahram Salehi Comparing Solo Versus Ensemble Convolutional Neural Networks for Wetland Classification Using Multi-Spectral Satellite Imagery |
topic_facet |
deep learning wetland mapping convolutional neural network satellite image classification ensemble learning Science Q |
description |
Wetlands are important ecosystems that are linked to climate change mitigation. As 25% of global wetlands are located in Canada, accurate and up-to-date wetland classification is of high importance, nationally and internationally. The advent of deep learning techniques has revolutionized the current use of machine learning algorithms to classify complex environments, specifically in remote sensing. In this paper, we explore the potential and possible limitations to be overcome regarding the use of ensemble deep learning techniques for complex wetland classification and discusses the potential and limitation of various solo convolutional neural networks (CNNs), including DenseNet, GoogLeNet, ShuffleNet, MobileNet, Xception, Inception-ResNet, ResNet18, and ResNet101 in three different study areas located in Newfoundland and Labrador, Canada (i.e., Avalon, Gros Morne, and Grand Falls). Moreover, to improve the classification accuracies of wetland classes of bog, fen, marsh, swamp, and shallow water, the results of the three best CNNs in each study area is fused using three supervised classifiers of random forest (RF), bagged tree (BTree), Bayesian optimized tree (BOT), and one unsupervised majority voting classifier. The results suggest that the ensemble models, in particular BTree, have a valuable role to play in the classification of wetland classes of bog, fen, marsh, swamp, and shallow water. The ensemble CNNs show an improvement of 9.63–19.04% in terms of mean producer’s accuracy compared to the solo CNNs, to recognize wetland classes in three different study areas. This research indicates a promising potential for integrating ensemble-based learning and deep learning for operational large area land cover, particularly complex wetland type classification. |
format |
Article in Journal/Newspaper |
author |
Ali Jamali Masoud Mahdianpari Brian Brisco Jean Granger Fariba Mohammadimanesh Bahram Salehi |
author_facet |
Ali Jamali Masoud Mahdianpari Brian Brisco Jean Granger Fariba Mohammadimanesh Bahram Salehi |
author_sort |
Ali Jamali |
title |
Comparing Solo Versus Ensemble Convolutional Neural Networks for Wetland Classification Using Multi-Spectral Satellite Imagery |
title_short |
Comparing Solo Versus Ensemble Convolutional Neural Networks for Wetland Classification Using Multi-Spectral Satellite Imagery |
title_full |
Comparing Solo Versus Ensemble Convolutional Neural Networks for Wetland Classification Using Multi-Spectral Satellite Imagery |
title_fullStr |
Comparing Solo Versus Ensemble Convolutional Neural Networks for Wetland Classification Using Multi-Spectral Satellite Imagery |
title_full_unstemmed |
Comparing Solo Versus Ensemble Convolutional Neural Networks for Wetland Classification Using Multi-Spectral Satellite Imagery |
title_sort |
comparing solo versus ensemble convolutional neural networks for wetland classification using multi-spectral satellite imagery |
publisher |
MDPI AG |
publishDate |
2021 |
url |
https://doi.org/10.3390/rs13112046 https://doaj.org/article/924bddcdfb254a38885592d6bedf0cae |
geographic |
Newfoundland Canada |
geographic_facet |
Newfoundland Canada |
genre |
Newfoundland |
genre_facet |
Newfoundland |
op_source |
Remote Sensing, Vol 13, Iss 2046, p 2046 (2021) |
op_relation |
https://www.mdpi.com/2072-4292/13/11/2046 https://doaj.org/toc/2072-4292 doi:10.3390/rs13112046 2072-4292 https://doaj.org/article/924bddcdfb254a38885592d6bedf0cae |
op_doi |
https://doi.org/10.3390/rs13112046 |
container_title |
Remote Sensing |
container_volume |
13 |
container_issue |
11 |
container_start_page |
2046 |
_version_ |
1766109552534618112 |