Repository logo
 

Image super-resolution with dense-sampling residual channel-spatial attention networks for multi-temporal remote sensing image classification

Simple item page
cam.issuedOnline2021-09-20
dc.contributor.authorZhu, Y
dc.contributor.authorGeiß, C
dc.contributor.authorSo, E
dc.contributor.orcidZhu, Yue [0000-0002-3154-9659]
dc.date.accessioned2021-10-26T23:31:08Z
dc.date.available2021-10-26T23:31:08Z
dc.date.issued2021
dc.description.abstractImage super-resolution (SR) techniques can benefit a wide range of applications in the remote sensing (RS) community, including image classification. This issue is particularly relevant for image classification on time series data, considering RS datasets that feature long temporal coverage generally have a limited spatial resolution. Recent advances in deep learning brought new opportunities for enhancing the spatial resolution of historic RS data. Numerous convolutional neural network (CNN)-based methods showed superior performance in terms of developing efficient end-to-end SR models for natural images. However, such models were rarely exploited for promoting image classification based on multispectral RS data. This paper proposes a novel CNNbased framework to enhance the spatial resolution of time series multispectral RS images. Thereby, the proposed SR model employs Residual Channel Attention Networks (RCAN) as a backbone structure, whereas based on this structure the proposed models uniquely integrate tailored channel-spatial attention and dense-sampling mechanisms for performance improvement. Subsequently, state-of-the-art CNN-based classifiers are incorporated to produce classification maps based on the enhanced time series data. The experiments proved that the proposed SR model can enable unambiguously better performance compared to RCAN and other (deep learning-based) SR techniques, especially in a domain adaptation context, i.e., leveraging Sentinel-2 images for generating SR Landsat images. Furthermore, the experimental results confirmed that the enhanced multi-temporal RS images can bring substantial improvement on fine-grained multi-temporal land use classification.
dc.identifier.doi10.17863/CAM.77373
dc.identifier.eissn1872-826X
dc.identifier.issn1569-8432
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/329930
dc.language.isoeng
dc.publisherElsevier BV
dc.publisher.urlhttp://dx.doi.org/10.1016/j.jag.2021.102543
dc.rightsAttribution-NonCommercial-NoDerivatives 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.subjectImage super-resolution
dc.subjectConvolutional neural networks
dc.subjectAttention mechanism
dc.subjectDense connection
dc.subjectMulti-temporal land use classification
dc.titleImage super-resolution with dense-sampling residual channel-spatial attention networks for multi-temporal remote sensing image classification
dc.typeArticle
dcterms.dateAccepted2021-09-12
prism.numberARTN 102543
prism.publicationDate2021
prism.publicationNameInternational Journal of Applied Earth Observation and Geoinformation
prism.volume104
pubs.funder-project-idEngineering and Physical Sciences Research Council (EP/P025234/1)
rioxxterms.licenseref.startdate2021-12-15
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserved
rioxxterms.typeJournal Article/Review
rioxxterms.versionVoR
rioxxterms.versionofrecord10.1016/j.jag.2021.102543

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Image super-resolution with dense-sampling residual channel-spatial.pdf
Size:
24.98 MB
Format:
Adobe Portable Document Format
Description:
Published version
Licence
https://creativecommons.org/licenses/by-nc-nd/4.0/
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
DepositLicenceAgreementv2.1.pdf
Size:
150.9 KB
Format:
Adobe Portable Document Format
Download

Collections

Cambridge University Research Outputs