MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice
High-resolution aerial photographs of Arctic region are a great source for different sea ice feature recognition, which are crucial to validate, tune and improve climate models. Melt ponds on the surface of melting Arctic sea ice are of particular interest as they are sensitive and valuable indicato...
Published in: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
---|---|
Main Authors: | , , , |
Language: | unknown |
Published: |
2022
|
Subjects: | |
Online Access: | https://doi.org/10.1109/JSTARS.2022.3213192 https://research.chalmers.se/en/publication/532704 |
id |
ftchalmersuniv:oai:research.chalmers.se:532704 |
---|---|
record_format |
openpolar |
spelling |
ftchalmersuniv:oai:research.chalmers.se:532704 2023-05-15T14:38:46+02:00 MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice Sudakow, Ivan Asari, Vijayan K. Liu, Ruixu Demchev, Denis 2022 text https://doi.org/10.1109/JSTARS.2022.3213192 https://research.chalmers.se/en/publication/532704 unknown http://dx.doi.org/10.1109/JSTARS.2022.3213192 https://research.chalmers.se/en/publication/532704 Climate Research Annotations Arctic Decoding Transformers melt ponds remote sensing Image segmentation complex system sea ice Data models swin transformer deep learning 2022 ftchalmersuniv https://doi.org/10.1109/JSTARS.2022.3213192 2022-12-11T07:19:55Z High-resolution aerial photographs of Arctic region are a great source for different sea ice feature recognition, which are crucial to validate, tune and improve climate models. Melt ponds on the surface of melting Arctic sea ice are of particular interest as they are sensitive and valuable indicators and are proxy to the processes in the Arctic climate system. Manual analysis of this remote sensing data is extremely difficult and time-consuming due to the complex shapes and unpredictable boundaries of the melt ponds, and that leads to the necessity for automatizing the processes. In this study, we propose a robust and efficient automatic method for melt pond region segmentation and boundary extraction from high-resolution aerial photographs. The proposed algorithm is based on a swin transformer U-Net in which we introduce novel cross-channel attention mechanisms into the decoder design. The framework operates with optical data and allows for classifying imagery into four classes: sea ice/snow, open water, melt pond, and submerged ice. We use aerial photographs collected during the Healy-Oden Trans Arctic Expedition (HO-TRAX) expedition over Arctic sea ice in the summer season of 2005 as test data. The experimental results show that the proposed method is suitable for precise automatic extraction of melt pond geometry and it can also be extended for other optical data sources that involve melt ponds. The approach has a promising potential to be used to analyze melt ponds' corresponding changes between years. Other/Unknown Material Arctic Sea ice Chalmers University of Technology: Chalmers research Arctic IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 15 8776 8784 |
institution |
Open Polar |
collection |
Chalmers University of Technology: Chalmers research |
op_collection_id |
ftchalmersuniv |
language |
unknown |
topic |
Climate Research Annotations Arctic Decoding Transformers melt ponds remote sensing Image segmentation complex system sea ice Data models swin transformer deep learning |
spellingShingle |
Climate Research Annotations Arctic Decoding Transformers melt ponds remote sensing Image segmentation complex system sea ice Data models swin transformer deep learning Sudakow, Ivan Asari, Vijayan K. Liu, Ruixu Demchev, Denis MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice |
topic_facet |
Climate Research Annotations Arctic Decoding Transformers melt ponds remote sensing Image segmentation complex system sea ice Data models swin transformer deep learning |
description |
High-resolution aerial photographs of Arctic region are a great source for different sea ice feature recognition, which are crucial to validate, tune and improve climate models. Melt ponds on the surface of melting Arctic sea ice are of particular interest as they are sensitive and valuable indicators and are proxy to the processes in the Arctic climate system. Manual analysis of this remote sensing data is extremely difficult and time-consuming due to the complex shapes and unpredictable boundaries of the melt ponds, and that leads to the necessity for automatizing the processes. In this study, we propose a robust and efficient automatic method for melt pond region segmentation and boundary extraction from high-resolution aerial photographs. The proposed algorithm is based on a swin transformer U-Net in which we introduce novel cross-channel attention mechanisms into the decoder design. The framework operates with optical data and allows for classifying imagery into four classes: sea ice/snow, open water, melt pond, and submerged ice. We use aerial photographs collected during the Healy-Oden Trans Arctic Expedition (HO-TRAX) expedition over Arctic sea ice in the summer season of 2005 as test data. The experimental results show that the proposed method is suitable for precise automatic extraction of melt pond geometry and it can also be extended for other optical data sources that involve melt ponds. The approach has a promising potential to be used to analyze melt ponds' corresponding changes between years. |
author |
Sudakow, Ivan Asari, Vijayan K. Liu, Ruixu Demchev, Denis |
author_facet |
Sudakow, Ivan Asari, Vijayan K. Liu, Ruixu Demchev, Denis |
author_sort |
Sudakow, Ivan |
title |
MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice |
title_short |
MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice |
title_full |
MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice |
title_fullStr |
MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice |
title_full_unstemmed |
MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice |
title_sort |
meltpondnet: a swin transformer u-net for detection of melt ponds on arctic sea ice |
publishDate |
2022 |
url |
https://doi.org/10.1109/JSTARS.2022.3213192 https://research.chalmers.se/en/publication/532704 |
geographic |
Arctic |
geographic_facet |
Arctic |
genre |
Arctic Sea ice |
genre_facet |
Arctic Sea ice |
op_relation |
http://dx.doi.org/10.1109/JSTARS.2022.3213192 https://research.chalmers.se/en/publication/532704 |
op_doi |
https://doi.org/10.1109/JSTARS.2022.3213192 |
container_title |
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
container_volume |
15 |
container_start_page |
8776 |
op_container_end_page |
8784 |
_version_ |
1766310797044088832 |