Physics-Based GAN With Iterative Refinement Unit for Hyperspectral and Multispectral Image Fusion

Hyperspectral image (HSI) fusion can effectively improve the spatial resolution of HSIs by integrating high-resolution multispectral images (MSIs). Considering the spatial and spectral degradation relationship between a fused image and input images, a physics-based GAN is proposed to fuse HSI and MS...

Full description

Bibliographic Details
Main Authors: Jiajun Xiao, Jie Li, Qiangqiang Yuan, Menghui Jiang, Liangpei Zhang
Format: Article
Language:English
Published: IEEE 2021-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/9435191/
_version_ 1818619377019256832
author Jiajun Xiao
Jie Li
Qiangqiang Yuan
Menghui Jiang
Liangpei Zhang
author_facet Jiajun Xiao
Jie Li
Qiangqiang Yuan
Menghui Jiang
Liangpei Zhang
author_sort Jiajun Xiao
collection DOAJ
description Hyperspectral image (HSI) fusion can effectively improve the spatial resolution of HSIs by integrating high-resolution multispectral images (MSIs). Considering the spatial and spectral degradation relationship between a fused image and input images, a physics-based GAN is proposed to fuse HSI and MSI. A physical model estimating degradation of image is introduced in the generator and in the discriminators. For the generator, a set of recursive modules including a physical degradation model and a multiscale residual channel attention fusion module integrate the spectral-spatial difference information between input images and estimated degradation images to restore the details of the fused image. Subsequently, the residual spatial attention fusion module is used to combine the results of all recursions to obtain the final reconstructed result. As for the discriminators, three networks with the final fused image, estimated LR HSI and estimated MSI as inputs share the same architecture. Finally, the loss function that contains adversarial losses and L1 losses of the fused image and estimated degradation images is used to optimize network parameters. The experimental results demonstrate that the proposed method outperforms state-of-the-art methods.
first_indexed 2024-12-16T17:36:31Z
format Article
id doaj.art-1a45fd5b14354d00a4106ed56a3c81e8
institution Directory Open Access Journal
issn 2151-1535
language English
last_indexed 2024-12-16T17:36:31Z
publishDate 2021-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj.art-1a45fd5b14354d00a4106ed56a3c81e82022-12-21T22:22:45ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352021-01-01146827684110.1109/JSTARS.2021.30757279435191Physics-Based GAN With Iterative Refinement Unit for Hyperspectral and Multispectral Image FusionJiajun Xiao0Jie Li1https://orcid.org/0000-0002-4063-9381Qiangqiang Yuan2https://orcid.org/0000-0001-7140-2224Menghui Jiang3Liangpei Zhang4https://orcid.org/0000-0001-6890-3650School of Geodesy and Geomatics, Wuhan University, Wuhan, ChinaSchool of Geodesy and Geomatics, Wuhan University, Wuhan, ChinaSchool of Geodesy and Geomatics, Wuhan University, Wuhan, ChinaSchool of Resource and Environmental Sciences, Wuhan University, Wuhan, ChinaState Key Laboratory of Information Engineering in Surveying, Mapping, and Remote Sensing, Wuhan University, Wuhan, ChinaHyperspectral image (HSI) fusion can effectively improve the spatial resolution of HSIs by integrating high-resolution multispectral images (MSIs). Considering the spatial and spectral degradation relationship between a fused image and input images, a physics-based GAN is proposed to fuse HSI and MSI. A physical model estimating degradation of image is introduced in the generator and in the discriminators. For the generator, a set of recursive modules including a physical degradation model and a multiscale residual channel attention fusion module integrate the spectral-spatial difference information between input images and estimated degradation images to restore the details of the fused image. Subsequently, the residual spatial attention fusion module is used to combine the results of all recursions to obtain the final reconstructed result. As for the discriminators, three networks with the final fused image, estimated LR HSI and estimated MSI as inputs share the same architecture. Finally, the loss function that contains adversarial losses and L1 losses of the fused image and estimated degradation images is used to optimize network parameters. The experimental results demonstrate that the proposed method outperforms state-of-the-art methods.https://ieeexplore.ieee.org/document/9435191/Attention moduleGANhyperspectral image (HSI)physics model
spellingShingle Jiajun Xiao
Jie Li
Qiangqiang Yuan
Menghui Jiang
Liangpei Zhang
Physics-Based GAN With Iterative Refinement Unit for Hyperspectral and Multispectral Image Fusion
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Attention module
GAN
hyperspectral image (HSI)
physics model
title Physics-Based GAN With Iterative Refinement Unit for Hyperspectral and Multispectral Image Fusion
title_full Physics-Based GAN With Iterative Refinement Unit for Hyperspectral and Multispectral Image Fusion
title_fullStr Physics-Based GAN With Iterative Refinement Unit for Hyperspectral and Multispectral Image Fusion
title_full_unstemmed Physics-Based GAN With Iterative Refinement Unit for Hyperspectral and Multispectral Image Fusion
title_short Physics-Based GAN With Iterative Refinement Unit for Hyperspectral and Multispectral Image Fusion
title_sort physics based gan with iterative refinement unit for hyperspectral and multispectral image fusion
topic Attention module
GAN
hyperspectral image (HSI)
physics model
url https://ieeexplore.ieee.org/document/9435191/
work_keys_str_mv AT jiajunxiao physicsbasedganwithiterativerefinementunitforhyperspectralandmultispectralimagefusion
AT jieli physicsbasedganwithiterativerefinementunitforhyperspectralandmultispectralimagefusion
AT qiangqiangyuan physicsbasedganwithiterativerefinementunitforhyperspectralandmultispectralimagefusion
AT menghuijiang physicsbasedganwithiterativerefinementunitforhyperspectralandmultispectralimagefusion
AT liangpeizhang physicsbasedganwithiterativerefinementunitforhyperspectralandmultispectralimagefusion