Super-Resolution of Radargrams with a Generative Deep Learning Model
Radar sounder (RS) profiles are essential for imaging the subsurface of planetary bodies and the Earth as they provide valuable geological insights. However, the limited availability of high-resolution radargrams poses challenges. This article proposes a novel method based on generative models to su...
Published in: | IEEE Transactions on Geoscience and Remote Sensing |
---|---|
Main Authors: | , , |
Other Authors: | , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
2024
|
Subjects: | |
Online Access: | https://hdl.handle.net/11582/345814 https://doi.org/10.1109/TGRS.2024.3378576 https://ieeexplore.ieee.org/document/10474029 |
id |
ftfbkssleriris:oai:cris.fbk.eu:11582/345814 |
---|---|
record_format |
openpolar |
spelling |
ftfbkssleriris:oai:cris.fbk.eu:11582/345814 2024-05-19T07:29:27+00:00 Super-Resolution of Radargrams with a Generative Deep Learning Model Elena Donini Lorenzo Bruzzone Francesca Bovolo Donini, Elena Bruzzone, Lorenzo Bovolo, Francesca 2024 ELETTRONICO https://hdl.handle.net/11582/345814 https://doi.org/10.1109/TGRS.2024.3378576 https://ieeexplore.ieee.org/document/10474029 eng eng volume:62 journal:IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING https://hdl.handle.net/11582/345814 doi:10.1109/TGRS.2024.3378576 https://ieeexplore.ieee.org/document/10474029 info:eu-repo/semantics/restrictedAccess info:eu-repo/semantics/article 2024 ftfbkssleriris https://doi.org/10.1109/TGRS.2024.3378576 2024-04-21T23:41:47Z Radar sounder (RS) profiles are essential for imaging the subsurface of planetary bodies and the Earth as they provide valuable geological insights. However, the limited availability of high-resolution radargrams poses challenges. This article proposes a novel method based on generative models to super-resolve radargrams. Our approach addresses the ill-posed and ill-conditioned nature of the super-resolution problem by training a neural network to learn the correlation between radargrams at different scales. The network learns a proxy for the mapping function between ambiguous low-resolution radargrams and more detailed high-resolution ones, considering the data’s geological and statistical properties. The mapping function enables the super-resolution of previously unseen low-resolution radargrams acquired in comparable conditions to those in the training and imaging similar underlying geology. To achieve this, we adopt a cycle generative adversarial network (CycleGAN), explicitly designed to match properties between low- and high-resolution radargrams, accounting for variations in dimensions and radiometric properties. Furthermore, we enhance the network performance by incorporating skip connections, a ResNet module, and attention mechanisms. The proposed method is validated using MCoRDS3 radargrams acquired in Greenland and Antarctica as high-resolution data. As low-resolution data, we used simulated radargrams representing what is expected by an Earth-orbiting low-resolution RS to have a controlled experiment. The results are evaluated qualitatively and quantitatively, focusing on the areas with reflections with complex shapes that may generate artifacts and unrealistic geological features. Article in Journal/Newspaper Antarc* Antarctica Greenland Fondazione Bruno Kessler: CINECA IRIS IEEE Transactions on Geoscience and Remote Sensing 62 1 17 |
institution |
Open Polar |
collection |
Fondazione Bruno Kessler: CINECA IRIS |
op_collection_id |
ftfbkssleriris |
language |
English |
description |
Radar sounder (RS) profiles are essential for imaging the subsurface of planetary bodies and the Earth as they provide valuable geological insights. However, the limited availability of high-resolution radargrams poses challenges. This article proposes a novel method based on generative models to super-resolve radargrams. Our approach addresses the ill-posed and ill-conditioned nature of the super-resolution problem by training a neural network to learn the correlation between radargrams at different scales. The network learns a proxy for the mapping function between ambiguous low-resolution radargrams and more detailed high-resolution ones, considering the data’s geological and statistical properties. The mapping function enables the super-resolution of previously unseen low-resolution radargrams acquired in comparable conditions to those in the training and imaging similar underlying geology. To achieve this, we adopt a cycle generative adversarial network (CycleGAN), explicitly designed to match properties between low- and high-resolution radargrams, accounting for variations in dimensions and radiometric properties. Furthermore, we enhance the network performance by incorporating skip connections, a ResNet module, and attention mechanisms. The proposed method is validated using MCoRDS3 radargrams acquired in Greenland and Antarctica as high-resolution data. As low-resolution data, we used simulated radargrams representing what is expected by an Earth-orbiting low-resolution RS to have a controlled experiment. The results are evaluated qualitatively and quantitatively, focusing on the areas with reflections with complex shapes that may generate artifacts and unrealistic geological features. |
author2 |
Donini, Elena Bruzzone, Lorenzo Bovolo, Francesca |
format |
Article in Journal/Newspaper |
author |
Elena Donini Lorenzo Bruzzone Francesca Bovolo |
spellingShingle |
Elena Donini Lorenzo Bruzzone Francesca Bovolo Super-Resolution of Radargrams with a Generative Deep Learning Model |
author_facet |
Elena Donini Lorenzo Bruzzone Francesca Bovolo |
author_sort |
Elena Donini |
title |
Super-Resolution of Radargrams with a Generative Deep Learning Model |
title_short |
Super-Resolution of Radargrams with a Generative Deep Learning Model |
title_full |
Super-Resolution of Radargrams with a Generative Deep Learning Model |
title_fullStr |
Super-Resolution of Radargrams with a Generative Deep Learning Model |
title_full_unstemmed |
Super-Resolution of Radargrams with a Generative Deep Learning Model |
title_sort |
super-resolution of radargrams with a generative deep learning model |
publishDate |
2024 |
url |
https://hdl.handle.net/11582/345814 https://doi.org/10.1109/TGRS.2024.3378576 https://ieeexplore.ieee.org/document/10474029 |
genre |
Antarc* Antarctica Greenland |
genre_facet |
Antarc* Antarctica Greenland |
op_relation |
volume:62 journal:IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING https://hdl.handle.net/11582/345814 doi:10.1109/TGRS.2024.3378576 https://ieeexplore.ieee.org/document/10474029 |
op_rights |
info:eu-repo/semantics/restrictedAccess |
op_doi |
https://doi.org/10.1109/TGRS.2024.3378576 |
container_title |
IEEE Transactions on Geoscience and Remote Sensing |
container_volume |
62 |
container_start_page |
1 |
op_container_end_page |
17 |
_version_ |
1799479382252716032 |