Consiglio Nazionale delle Ricerche

Tipo di prodottoArticolo in rivista
TitoloA New Variational Approach Based on Proximal Deep Injection and Gradient Intensity Similarity for Spatio-Spectral Image Fusion
Anno di pubblicazione2020
Formato
  • Elettronico
  • Cartaceo
Autore/iWu, Zhong-Cheng; Huang, Ting-Zhu; Deng, Liang-Jian; Vivone, Gemine; Miao, Jia-Qing; Hu, Jin-Fan; Zhao, Xi-Le
Affiliazioni autoria.School of Mathematical Sciences, University of Electronic Science and Technology of China, Chengdu, 611731, China b.CNR-IMAA, National Research Council - Institute of Methodologies for Environmental Analysis, Tito Scalo, 85050, Italy c.School of Computer Science, and Technology, University of Southwest Minzu of China, Chengdu, 610041, China
Autori CNR e affiliazioni
  • GEMINE VIVONE
Lingua/e
  • inglese
AbstractPansharpening is a very debated spatio-spectral fusion problem. It refers to the fusion of a high spatial resolution panchromatic image with a lower spatial but higher spectral resolution multispectral image in order to obtain an image with high resolution in both the domains. In this article, we propose a novel variational optimization-based (VO) approach to address this issue incorporating the outcome of a deep convolutional neural network (DCNN). This solution can take advantages of both the paradigms. On one hand, higher performance can be expected introducing machine learning (ML) methods based on the training by examples philosophy into VO approaches. On other hand, the combination of VO techniques with DCNNs can aid the generalization ability of these latter. In particular, we formulate a l(2) -based proximal deep injection term to evaluate the distance between the DCNN outcome, and the desired high spatial resolution multispectral image. This represents the regularization term for our VO model. Furthermore, a new data fitting term measuring the spatial fidelity is proposed. Finally, the proposed convex VO problem is efficiently solved by exploiting the framework of the alternating direction method of multipliers(ADMM), thus guaranteeing the convergence of the algorithm. Extensive experiments both on simulated, and real datasets demonstrate that the proposed approach can outperform state-of-the-art spatio-spectral fusion methods, even showing a significant generalization ability. Please find the project page at https: //liangjiandeng.github.io/ Projects_Res/DMPIF_2020jstars.html.
Lingua abstractinglese
Altro abstract-
Lingua altro abstract-
Pagine da6277
Pagine a6290
Pagine totali14
RivistaIEEE journal of selected topics in applied earth observations and remote sensing (Print)
Attiva dal 2008
Editore: IEEE, - Piscataway, N.J.
Paese di pubblicazione: Stati Uniti d'America
Lingua: inglese
ISSN: 1939-1404
Titolo chiave: IEEE journal of selected topics in applied earth observations and remote sensing (Print)
Titolo proprio: IEEE journal of selected topics in applied earth observations and remote sensing. (Print)
Titolo abbreviato: IEEE j. sel. top. appl. earth obs. remote sens. (Print)
Titoli alternativi:
  • Institute of Electrical and Electronic Engineers journal of selected topics in applied earth observations and remote sensing (Print)
  • Journal of selected topics in applied earth observations and remote sensing (Print)
  • J-STARS (Print)
Numero volume della rivista13
Fascicolo della rivista-
DOI10.1109/JSTARS.2020.3030129
Verificato da refereeSì: Internazionale
Stato della pubblicazionePublished version
Indicizzazione (in banche dati controllate)
  • ISI Web of Science (WOS) (Codice:000583704200001)
  • ISI Web of Science (WOS) (Codice:2-s2.0-85095684713)
Parole chiaveSpatial resolution, Optimization, Training, Image fusion, Convolutional neural networks, Data models, Deep convolutional neural networks (DCNN), dynamic gradient sparsity, gradient intensity similarity, image fusion, pansharpening, remote sensing, variational approaches
Link (URL, URI)https://ieeexplore.ieee.org/document/9220783
Titolo parallelo-
Licenza-
Scadenza embargo-
Data di accettazione-
Note/Altre informazioni-
Strutture CNR
  • IMAA — Istituto di metodologie per l'analisi ambientale
Moduli/Attività/Sottoprogetti CNR-
Progetti Europei-
Allegati
A New Variational Approach Based on Proximal Deep Injection and Gradient Intensity Similarity for Spatio-Spectral Image Fusion (documento privato )
Tipo documento: application/pdf