HyperGAN: A Hyperspectral Image Fusion Approach Based on Generative Adversarial Networks

The objective of hyperspectral pansharpening is to fuse low-resolution hyperspectral images (LR-HSI) with corresponding panchromatic (PAN) images to generate high-resolution hyperspectral images (HR-HSI). Despite advancements in hyperspectral (HS) pansharpening using deep learning, the rich spectral...

Full description

Saved in:
Bibliographic Details
Main Authors: Jing Wang, Xu Zhu, Linhai Jing, Yunwei Tang, Hui Li, Zhengqing Xiao, Haifeng Ding
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/16/23/4389
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846123939746545664
author Jing Wang
Xu Zhu
Linhai Jing
Yunwei Tang
Hui Li
Zhengqing Xiao
Haifeng Ding
author_facet Jing Wang
Xu Zhu
Linhai Jing
Yunwei Tang
Hui Li
Zhengqing Xiao
Haifeng Ding
author_sort Jing Wang
collection DOAJ
description The objective of hyperspectral pansharpening is to fuse low-resolution hyperspectral images (LR-HSI) with corresponding panchromatic (PAN) images to generate high-resolution hyperspectral images (HR-HSI). Despite advancements in hyperspectral (HS) pansharpening using deep learning, the rich spectral details and large data volume of HS images place higher demands on models for effective spectral extraction and processing. In this paper, we present HyperGAN, a hyperspectral image fusion approach based on Generative Adversarial Networks. Unlike previous methods that deepen the network to capture spectral information, HyperGAN widens the structure with a Wide Block for multi-scale learning, effectively capturing global and local details from upsampled HSI and PAN images. While LR-HSI provides rich spectral data, PAN images offer spatial information. We introduce the Efficient Spatial and Channel Attention Module (ESCA) to integrate these features and add an energy-based discriminator to enhance model performance by learning directly from the Ground Truth (GT), improving fused image quality. We validated our method on various scenes, including the Pavia Center, Eastern Tianshan, and Chikusei. Results show that HyperGAN outperforms state-of-the-art methods in visual and quantitative evaluations.
format Article
id doaj-art-c43089af11aa4560bfbf900ab27a8856
institution Kabale University
issn 2072-4292
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-c43089af11aa4560bfbf900ab27a88562024-12-13T16:30:40ZengMDPI AGRemote Sensing2072-42922024-11-011623438910.3390/rs16234389HyperGAN: A Hyperspectral Image Fusion Approach Based on Generative Adversarial NetworksJing Wang0Xu Zhu1Linhai Jing2Yunwei Tang3Hui Li4Zhengqing Xiao5Haifeng Ding6College of Mathematics and System Science, Xinjiang University, Urumqi 830017, ChinaCollege of Mathematics and System Science, Xinjiang University, Urumqi 830017, ChinaSchool of Information Engineering, China University of Geosciences (Beijing), Beijing 100083, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaCollege of Mathematics and System Science, Xinjiang University, Urumqi 830017, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaThe objective of hyperspectral pansharpening is to fuse low-resolution hyperspectral images (LR-HSI) with corresponding panchromatic (PAN) images to generate high-resolution hyperspectral images (HR-HSI). Despite advancements in hyperspectral (HS) pansharpening using deep learning, the rich spectral details and large data volume of HS images place higher demands on models for effective spectral extraction and processing. In this paper, we present HyperGAN, a hyperspectral image fusion approach based on Generative Adversarial Networks. Unlike previous methods that deepen the network to capture spectral information, HyperGAN widens the structure with a Wide Block for multi-scale learning, effectively capturing global and local details from upsampled HSI and PAN images. While LR-HSI provides rich spectral data, PAN images offer spatial information. We introduce the Efficient Spatial and Channel Attention Module (ESCA) to integrate these features and add an energy-based discriminator to enhance model performance by learning directly from the Ground Truth (GT), improving fused image quality. We validated our method on various scenes, including the Pavia Center, Eastern Tianshan, and Chikusei. Results show that HyperGAN outperforms state-of-the-art methods in visual and quantitative evaluations.https://www.mdpi.com/2072-4292/16/23/4389generative adversarial networkshyperspectral pansharpeningattentionenergy
spellingShingle Jing Wang
Xu Zhu
Linhai Jing
Yunwei Tang
Hui Li
Zhengqing Xiao
Haifeng Ding
HyperGAN: A Hyperspectral Image Fusion Approach Based on Generative Adversarial Networks
Remote Sensing
generative adversarial networks
hyperspectral pansharpening
attention
energy
title HyperGAN: A Hyperspectral Image Fusion Approach Based on Generative Adversarial Networks
title_full HyperGAN: A Hyperspectral Image Fusion Approach Based on Generative Adversarial Networks
title_fullStr HyperGAN: A Hyperspectral Image Fusion Approach Based on Generative Adversarial Networks
title_full_unstemmed HyperGAN: A Hyperspectral Image Fusion Approach Based on Generative Adversarial Networks
title_short HyperGAN: A Hyperspectral Image Fusion Approach Based on Generative Adversarial Networks
title_sort hypergan a hyperspectral image fusion approach based on generative adversarial networks
topic generative adversarial networks
hyperspectral pansharpening
attention
energy
url https://www.mdpi.com/2072-4292/16/23/4389
work_keys_str_mv AT jingwang hyperganahyperspectralimagefusionapproachbasedongenerativeadversarialnetworks
AT xuzhu hyperganahyperspectralimagefusionapproachbasedongenerativeadversarialnetworks
AT linhaijing hyperganahyperspectralimagefusionapproachbasedongenerativeadversarialnetworks
AT yunweitang hyperganahyperspectralimagefusionapproachbasedongenerativeadversarialnetworks
AT huili hyperganahyperspectralimagefusionapproachbasedongenerativeadversarialnetworks
AT zhengqingxiao hyperganahyperspectralimagefusionapproachbasedongenerativeadversarialnetworks
AT haifengding hyperganahyperspectralimagefusionapproachbasedongenerativeadversarialnetworks