Multiscale Hyperspectral Pansharpening Network Based on Dual Pyramid and Transformer

Hyperspectral pansharpening is to fuse a high spatial resolution panchromatic image (PAN) with a low spatial resolution hyperspectral image (LR-HSI) and generate high resolution hyperspectral image (HR-HSI). However, most existing deep learning-based pansharpening methods have some issues, such as s...

Full description

Saved in:
Bibliographic Details
Main Authors: Hengyou Wang, Jie Zhang, Lian-Zhi Huo
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10545580/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846169467256569856
author Hengyou Wang
Jie Zhang
Lian-Zhi Huo
author_facet Hengyou Wang
Jie Zhang
Lian-Zhi Huo
author_sort Hengyou Wang
collection DOAJ
description Hyperspectral pansharpening is to fuse a high spatial resolution panchromatic image (PAN) with a low spatial resolution hyperspectral image (LR-HSI) and generate high resolution hyperspectral image (HR-HSI). However, most existing deep learning-based pansharpening methods have some issues, such as spectral distortion and insufficient spatial texture enhancement. In this work, we propose a novel multiscale pansharpening network based on the Dual Gaussian-Laplacian Pyramid (DGLP) and Transformer, named MDTP-Net. Specifically, the DGLP module is designed to obtain feature maps at multilevel scales, which effectively learn global spectral information and spatial detail texture information. Then, we design a corresponding Transformer module for each scale feature and utilize the multihead attention mechanism to guide the extraction of spatial information from LR-HSI and PAN images. This enhances the stability of pansharpening and improves the fusion of spectral with spatial information across feature spaces. In addition, the feature extractors are inserted to connect DGLP and Transformer, making the spatial feature map smoother and richer in channel and texture features. The feature fusion and multiscale feature connection blocks are used to connect multiscale information together to generate HR-HSI images with more comprehensive spatial and spectral features. Finally, extensive experiments on three classic hyperspectral datasets are conducted. The experimental results demonstrate that our proposed MDTP-Net outperforms conventional methods and existing deep learning-based methods.
format Article
id doaj-art-28e0d874cfb24235b4850b84aeb7809d
institution Kabale University
issn 1939-1404
2151-1535
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-28e0d874cfb24235b4850b84aeb7809d2024-11-13T00:00:15ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352024-01-0117197861979710.1109/JSTARS.2024.340828010545580Multiscale Hyperspectral Pansharpening Network Based on Dual Pyramid and TransformerHengyou Wang0https://orcid.org/0000-0001-6693-0161Jie Zhang1https://orcid.org/0009-0006-7029-6830Lian-Zhi Huo2https://orcid.org/0000-0001-6705-6453School of Science, Beijing University of Civil Engineering and Architecture, Beijing, ChinaSchool of Science, Beijing University of Civil Engineering and Architecture, Beijing, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaHyperspectral pansharpening is to fuse a high spatial resolution panchromatic image (PAN) with a low spatial resolution hyperspectral image (LR-HSI) and generate high resolution hyperspectral image (HR-HSI). However, most existing deep learning-based pansharpening methods have some issues, such as spectral distortion and insufficient spatial texture enhancement. In this work, we propose a novel multiscale pansharpening network based on the Dual Gaussian-Laplacian Pyramid (DGLP) and Transformer, named MDTP-Net. Specifically, the DGLP module is designed to obtain feature maps at multilevel scales, which effectively learn global spectral information and spatial detail texture information. Then, we design a corresponding Transformer module for each scale feature and utilize the multihead attention mechanism to guide the extraction of spatial information from LR-HSI and PAN images. This enhances the stability of pansharpening and improves the fusion of spectral with spatial information across feature spaces. In addition, the feature extractors are inserted to connect DGLP and Transformer, making the spatial feature map smoother and richer in channel and texture features. The feature fusion and multiscale feature connection blocks are used to connect multiscale information together to generate HR-HSI images with more comprehensive spatial and spectral features. Finally, extensive experiments on three classic hyperspectral datasets are conducted. The experimental results demonstrate that our proposed MDTP-Net outperforms conventional methods and existing deep learning-based methods.https://ieeexplore.ieee.org/document/10545580/Dual Gaussian-Laplacian pyramid (DGLP)hyperspectral pansharpeningmultiscale fusionTransformer
spellingShingle Hengyou Wang
Jie Zhang
Lian-Zhi Huo
Multiscale Hyperspectral Pansharpening Network Based on Dual Pyramid and Transformer
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Dual Gaussian-Laplacian pyramid (DGLP)
hyperspectral pansharpening
multiscale fusion
Transformer
title Multiscale Hyperspectral Pansharpening Network Based on Dual Pyramid and Transformer
title_full Multiscale Hyperspectral Pansharpening Network Based on Dual Pyramid and Transformer
title_fullStr Multiscale Hyperspectral Pansharpening Network Based on Dual Pyramid and Transformer
title_full_unstemmed Multiscale Hyperspectral Pansharpening Network Based on Dual Pyramid and Transformer
title_short Multiscale Hyperspectral Pansharpening Network Based on Dual Pyramid and Transformer
title_sort multiscale hyperspectral pansharpening network based on dual pyramid and transformer
topic Dual Gaussian-Laplacian pyramid (DGLP)
hyperspectral pansharpening
multiscale fusion
Transformer
url https://ieeexplore.ieee.org/document/10545580/
work_keys_str_mv AT hengyouwang multiscalehyperspectralpansharpeningnetworkbasedondualpyramidandtransformer
AT jiezhang multiscalehyperspectralpansharpeningnetworkbasedondualpyramidandtransformer
AT lianzhihuo multiscalehyperspectralpansharpeningnetworkbasedondualpyramidandtransformer