Sea Ice Concentration Estimation during Freeze-Up from SAR Imagery Using a Convolutional Neural Network
In this study, a convolutional neural network (CNN) is used to estimate sea ice concentration using synthetic aperture radar (SAR) scenes acquired during freeze-up in the Gulf of St. Lawrence on the east coast of Canada. The ice concentration estimates from the CNN are compared to those from a neura...
Published in: | Remote Sensing |
---|---|
Main Authors: | , , |
Format: | Text |
Language: | English |
Published: |
Multidisciplinary Digital Publishing Institute
2017
|
Subjects: | |
Online Access: | https://doi.org/10.3390/rs9050408 |
id |
ftmdpi:oai:mdpi.com:/2072-4292/9/5/408/ |
---|---|
record_format |
openpolar |
spelling |
ftmdpi:oai:mdpi.com:/2072-4292/9/5/408/ 2023-08-20T04:09:43+02:00 Sea Ice Concentration Estimation during Freeze-Up from SAR Imagery Using a Convolutional Neural Network Lei Wang K. Scott David Clausi agris 2017-04-26 application/pdf https://doi.org/10.3390/rs9050408 EN eng Multidisciplinary Digital Publishing Institute Remote Sensing Image Processing https://dx.doi.org/10.3390/rs9050408 https://creativecommons.org/licenses/by/4.0/ Remote Sensing; Volume 9; Issue 5; Pages: 408 ice concentration SAR imagery convolutional neural network Text 2017 ftmdpi https://doi.org/10.3390/rs9050408 2023-07-31T21:06:14Z In this study, a convolutional neural network (CNN) is used to estimate sea ice concentration using synthetic aperture radar (SAR) scenes acquired during freeze-up in the Gulf of St. Lawrence on the east coast of Canada. The ice concentration estimates from the CNN are compared to those from a neural network (multi-layer perceptron or MLP) that uses hand-crafted features as input and a single layer of hidden nodes. The CNN is found to be less sensitive to pixel level details than the MLP and produces ice concentration that is less noisy and in closer agreement with that from image analysis charts. This is due to the multi-layer (deep) structure of the CNN, which enables abstract image features to be learned. The CNN ice concentration is also compared with ice concentration estimated from passive microwave brightness temperature data using the ARTIST sea ice (ASI) algorithm. The bias and RMS of the difference between the ice concentration from the CNN and that from image analysis charts is reduced as compared to that from either the MLP or ASI algorithm. Additional results demonstrate the impact of varying the input patch size, varying the number of CNN layers, and including the incidence angle as an additional input. Text Sea ice MDPI Open Access Publishing Canada Remote Sensing 9 5 408 |
institution |
Open Polar |
collection |
MDPI Open Access Publishing |
op_collection_id |
ftmdpi |
language |
English |
topic |
ice concentration SAR imagery convolutional neural network |
spellingShingle |
ice concentration SAR imagery convolutional neural network Lei Wang K. Scott David Clausi Sea Ice Concentration Estimation during Freeze-Up from SAR Imagery Using a Convolutional Neural Network |
topic_facet |
ice concentration SAR imagery convolutional neural network |
description |
In this study, a convolutional neural network (CNN) is used to estimate sea ice concentration using synthetic aperture radar (SAR) scenes acquired during freeze-up in the Gulf of St. Lawrence on the east coast of Canada. The ice concentration estimates from the CNN are compared to those from a neural network (multi-layer perceptron or MLP) that uses hand-crafted features as input and a single layer of hidden nodes. The CNN is found to be less sensitive to pixel level details than the MLP and produces ice concentration that is less noisy and in closer agreement with that from image analysis charts. This is due to the multi-layer (deep) structure of the CNN, which enables abstract image features to be learned. The CNN ice concentration is also compared with ice concentration estimated from passive microwave brightness temperature data using the ARTIST sea ice (ASI) algorithm. The bias and RMS of the difference between the ice concentration from the CNN and that from image analysis charts is reduced as compared to that from either the MLP or ASI algorithm. Additional results demonstrate the impact of varying the input patch size, varying the number of CNN layers, and including the incidence angle as an additional input. |
format |
Text |
author |
Lei Wang K. Scott David Clausi |
author_facet |
Lei Wang K. Scott David Clausi |
author_sort |
Lei Wang |
title |
Sea Ice Concentration Estimation during Freeze-Up from SAR Imagery Using a Convolutional Neural Network |
title_short |
Sea Ice Concentration Estimation during Freeze-Up from SAR Imagery Using a Convolutional Neural Network |
title_full |
Sea Ice Concentration Estimation during Freeze-Up from SAR Imagery Using a Convolutional Neural Network |
title_fullStr |
Sea Ice Concentration Estimation during Freeze-Up from SAR Imagery Using a Convolutional Neural Network |
title_full_unstemmed |
Sea Ice Concentration Estimation during Freeze-Up from SAR Imagery Using a Convolutional Neural Network |
title_sort |
sea ice concentration estimation during freeze-up from sar imagery using a convolutional neural network |
publisher |
Multidisciplinary Digital Publishing Institute |
publishDate |
2017 |
url |
https://doi.org/10.3390/rs9050408 |
op_coverage |
agris |
geographic |
Canada |
geographic_facet |
Canada |
genre |
Sea ice |
genre_facet |
Sea ice |
op_source |
Remote Sensing; Volume 9; Issue 5; Pages: 408 |
op_relation |
Remote Sensing Image Processing https://dx.doi.org/10.3390/rs9050408 |
op_rights |
https://creativecommons.org/licenses/by/4.0/ |
op_doi |
https://doi.org/10.3390/rs9050408 |
container_title |
Remote Sensing |
container_volume |
9 |
container_issue |
5 |
container_start_page |
408 |
_version_ |
1774723354136150016 |