Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder Network
In this paper, we propose a novel hyperspectral image superresolution method based on superpixel spectral unmixing using a coupled encoder-decoder network. The hyperspectral image and multispectral images are fused to generate high-resolution hyperspectral images through the spectral unmixing framew...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Hindawi Limited
2020-01-01
|
Series: | Journal of Sensors |
Online Access: | http://dx.doi.org/10.1155/2020/8886178 |
id |
doaj-f2e2ea9b187744a0957fc0fbb2c15f4e |
---|---|
record_format |
Article |
spelling |
doaj-f2e2ea9b187744a0957fc0fbb2c15f4e2020-11-25T03:58:20ZengHindawi LimitedJournal of Sensors1687-725X1687-72682020-01-01202010.1155/2020/88861788886178Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder NetworkShao-lei Zhang0Guang-yuan Fu1Hong-qiao Wang2Yu-qing Zhao3Xi’an Research Institute of Hi-Tech, Hongqing Town, Xi’an, ChinaXi’an Research Institute of Hi-Tech, Hongqing Town, Xi’an, ChinaXi’an Research Institute of Hi-Tech, Hongqing Town, Xi’an, ChinaXi’an Research Institute of Hi-Tech, Hongqing Town, Xi’an, ChinaIn this paper, we propose a novel hyperspectral image superresolution method based on superpixel spectral unmixing using a coupled encoder-decoder network. The hyperspectral image and multispectral images are fused to generate high-resolution hyperspectral images through the spectral unmixing framework with low-rank constraint. Specifically, the endmember and abundance information is extracted via a coupled encoder-decoder network integrating the priori for unmixing. The coupled network consists of two encoders and one shared decoder, where spectral information is preserved through the encoder. The multispectral image is clustered into superpixels to explore self-similarity, and then, the superpixels are unmixed to obtain an abundance matrix. By imposing a low-rank constraint on the abundance matrix, we further improve the superresolution performance. Experiments on the CAVE and Harvard datasets indicate that our superresolution method outperforms the other compared methods in terms of quantitative evaluation and visual quality.http://dx.doi.org/10.1155/2020/8886178 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Shao-lei Zhang Guang-yuan Fu Hong-qiao Wang Yu-qing Zhao |
spellingShingle |
Shao-lei Zhang Guang-yuan Fu Hong-qiao Wang Yu-qing Zhao Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder Network Journal of Sensors |
author_facet |
Shao-lei Zhang Guang-yuan Fu Hong-qiao Wang Yu-qing Zhao |
author_sort |
Shao-lei Zhang |
title |
Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder Network |
title_short |
Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder Network |
title_full |
Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder Network |
title_fullStr |
Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder Network |
title_full_unstemmed |
Superpixel Spectral Unmixing for Hyperspectral Image Superresolution Using a Coupled Encoder-Decoder Network |
title_sort |
superpixel spectral unmixing for hyperspectral image superresolution using a coupled encoder-decoder network |
publisher |
Hindawi Limited |
series |
Journal of Sensors |
issn |
1687-725X 1687-7268 |
publishDate |
2020-01-01 |
description |
In this paper, we propose a novel hyperspectral image superresolution method based on superpixel spectral unmixing using a coupled encoder-decoder network. The hyperspectral image and multispectral images are fused to generate high-resolution hyperspectral images through the spectral unmixing framework with low-rank constraint. Specifically, the endmember and abundance information is extracted via a coupled encoder-decoder network integrating the priori for unmixing. The coupled network consists of two encoders and one shared decoder, where spectral information is preserved through the encoder. The multispectral image is clustered into superpixels to explore self-similarity, and then, the superpixels are unmixed to obtain an abundance matrix. By imposing a low-rank constraint on the abundance matrix, we further improve the superresolution performance. Experiments on the CAVE and Harvard datasets indicate that our superresolution method outperforms the other compared methods in terms of quantitative evaluation and visual quality. |
url |
http://dx.doi.org/10.1155/2020/8886178 |
work_keys_str_mv |
AT shaoleizhang superpixelspectralunmixingforhyperspectralimagesuperresolutionusingacoupledencoderdecodernetwork AT guangyuanfu superpixelspectralunmixingforhyperspectralimagesuperresolutionusingacoupledencoderdecodernetwork AT hongqiaowang superpixelspectralunmixingforhyperspectralimagesuperresolutionusingacoupledencoderdecodernetwork AT yuqingzhao superpixelspectralunmixingforhyperspectralimagesuperresolutionusingacoupledencoderdecodernetwork |
_version_ |
1715076033878687744 |