A Robust Multispectral Reconstruction Network from RGB Images Trained by Diverse Satellite Data and Application in Classification and Detection Tasks

Multispectral images contain richer spectral signatures than easily available RGB images, for which they are promising to contribute to information perception. However, the relatively high cost of multispectral sensors and lower spatial resolution limit the widespread application of multispectral da...

Full description

Saved in:
Bibliographic Details
Main Authors: Xiaoning Zhang, Zhaoyang Peng, Yifei Wang, Fan Ye, Tengying Fu, Hu Zhang
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/17/11/1901
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849722466911387648
author Xiaoning Zhang
Zhaoyang Peng
Yifei Wang
Fan Ye
Tengying Fu
Hu Zhang
author_facet Xiaoning Zhang
Zhaoyang Peng
Yifei Wang
Fan Ye
Tengying Fu
Hu Zhang
author_sort Xiaoning Zhang
collection DOAJ
description Multispectral images contain richer spectral signatures than easily available RGB images, for which they are promising to contribute to information perception. However, the relatively high cost of multispectral sensors and lower spatial resolution limit the widespread application of multispectral data, and existing reconstruction algorithms suffer from a lack of diverse training datasets and insufficient reconstruction accuracy. In response to these issues, this paper proposes a novel and robust multispectral reconstruction network from low-cost natural color RGB images based on free available satellite images with various land cover types. First, to supplement paired natural color RGB and multispectral images, the Houston hyperspectral dataset was used to train a convolutional neural network Model-TN for generating natural color RGB images from true color images combining CIE standard colorimetric system theory. Then, the EuroSAT multispectral satellite images for eight land cover types were selected to produce natural RGB using Model-TN as training image pairs, which were input into a residual network integrating channel attention mechanisms to train the multispectral images reconstruction model, Model-NM. Finally, the feasibility of the reconstructed multispectral images is verified through image classification and target detection. There is a small mean relative absolute error value of 0.0081 for generating natural color RGB images, which is 0.0397 for reconstructing multispectral images. Compared to RGB images, the accuracies of classification and detection using reconstructed multispectral images have improved by 16.67% and 3.09%, respectively. This study further reveals the potential of multispectral image reconstruction from natural color RGB images and its effectiveness in target detection, which promotes low-cost visual perception of intelligent unmanned systems.
format Article
id doaj-art-58f2585218b94e578ebd2d9696ccd85c
institution DOAJ
issn 2072-4292
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-58f2585218b94e578ebd2d9696ccd85c2025-08-20T03:11:20ZengMDPI AGRemote Sensing2072-42922025-05-011711190110.3390/rs17111901A Robust Multispectral Reconstruction Network from RGB Images Trained by Diverse Satellite Data and Application in Classification and Detection TasksXiaoning Zhang0Zhaoyang Peng1Yifei Wang2Fan Ye3Tengying Fu4Hu Zhang5School of Mechatronical Engineering, Beijing Institute of Technology, Beijing 100081, ChinaSchool of Mechatronical Engineering, Beijing Institute of Technology, Beijing 100081, ChinaSchool of Mechatronical Engineering, Beijing Institute of Technology, Beijing 100081, ChinaSchool of Mechatronical Engineering, Beijing Institute of Technology, Beijing 100081, ChinaSchool of Mechatronical Engineering, Beijing Institute of Technology, Beijing 100081, ChinaFaculty of Geographical, Tianjin Normal University, Tianjin 300387, ChinaMultispectral images contain richer spectral signatures than easily available RGB images, for which they are promising to contribute to information perception. However, the relatively high cost of multispectral sensors and lower spatial resolution limit the widespread application of multispectral data, and existing reconstruction algorithms suffer from a lack of diverse training datasets and insufficient reconstruction accuracy. In response to these issues, this paper proposes a novel and robust multispectral reconstruction network from low-cost natural color RGB images based on free available satellite images with various land cover types. First, to supplement paired natural color RGB and multispectral images, the Houston hyperspectral dataset was used to train a convolutional neural network Model-TN for generating natural color RGB images from true color images combining CIE standard colorimetric system theory. Then, the EuroSAT multispectral satellite images for eight land cover types were selected to produce natural RGB using Model-TN as training image pairs, which were input into a residual network integrating channel attention mechanisms to train the multispectral images reconstruction model, Model-NM. Finally, the feasibility of the reconstructed multispectral images is verified through image classification and target detection. There is a small mean relative absolute error value of 0.0081 for generating natural color RGB images, which is 0.0397 for reconstructing multispectral images. Compared to RGB images, the accuracies of classification and detection using reconstructed multispectral images have improved by 16.67% and 3.09%, respectively. This study further reveals the potential of multispectral image reconstruction from natural color RGB images and its effectiveness in target detection, which promotes low-cost visual perception of intelligent unmanned systems.https://www.mdpi.com/2072-4292/17/11/1901multispectral reconstructionresidual networkchannel attentionimage classificationtarget detection
spellingShingle Xiaoning Zhang
Zhaoyang Peng
Yifei Wang
Fan Ye
Tengying Fu
Hu Zhang
A Robust Multispectral Reconstruction Network from RGB Images Trained by Diverse Satellite Data and Application in Classification and Detection Tasks
Remote Sensing
multispectral reconstruction
residual network
channel attention
image classification
target detection
title A Robust Multispectral Reconstruction Network from RGB Images Trained by Diverse Satellite Data and Application in Classification and Detection Tasks
title_full A Robust Multispectral Reconstruction Network from RGB Images Trained by Diverse Satellite Data and Application in Classification and Detection Tasks
title_fullStr A Robust Multispectral Reconstruction Network from RGB Images Trained by Diverse Satellite Data and Application in Classification and Detection Tasks
title_full_unstemmed A Robust Multispectral Reconstruction Network from RGB Images Trained by Diverse Satellite Data and Application in Classification and Detection Tasks
title_short A Robust Multispectral Reconstruction Network from RGB Images Trained by Diverse Satellite Data and Application in Classification and Detection Tasks
title_sort robust multispectral reconstruction network from rgb images trained by diverse satellite data and application in classification and detection tasks
topic multispectral reconstruction
residual network
channel attention
image classification
target detection
url https://www.mdpi.com/2072-4292/17/11/1901
work_keys_str_mv AT xiaoningzhang arobustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT zhaoyangpeng arobustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT yifeiwang arobustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT fanye arobustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT tengyingfu arobustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT huzhang arobustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT xiaoningzhang robustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT zhaoyangpeng robustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT yifeiwang robustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT fanye robustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT tengyingfu robustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks
AT huzhang robustmultispectralreconstructionnetworkfromrgbimagestrainedbydiversesatellitedataandapplicationinclassificationanddetectiontasks