Pansharpening With Structure and Coefficient-Driven Structure Tensor Total Variation Priors
IEEE Transactions on Geoscience and Remote Sensing · 2026 · 64 · 1-14
DOI: 10.1109/tgrs.2026.3669775 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Geoscience and Remote Sensing, 2026. Verfasst von Pengfei Liu und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 57 Literaturangaben. Gefördert von: National Natural Science Foundation of China, China Postdoctoral Science Foundation, Frontier Technologies Research and Development Program of Jiangsu.
Abstract (Originalsprache)
In this article, we propose an innovative pansharpening method with structure and coefficient-driven structure tensor total variation (SCSTV) priors, which can obtain high-resolution multispectral (HRMS) images through the fusion of low-resolution multispectral (LRMS) and panchromatic (PAN) images, while effectively balancing the preservation of spatial structure and spectral information. On the one hand, based on various important structures in the HRMS image (i.e., edges, corners, and smooth regions), we employ the structure tensor for accurately modeling these desirable spatial structures in the HRMS image, and construct a structure-driven structure tensor total variation (STV) constraint for the HRMS image. On the other hand, based on the recently used pixelwise multiplicative coefficient model between HRMS and PAN images for spatial correlation modeling, we aim to apply the structure tensor descriptor for accurately modeling structures of the adaptive coefficient image (ACI) which also contains various important structures, such as edges, corners, and smooth regions, and construct a novel spatial fidelity term with the coefficient-driven STV prior, which is recently the first work to model the ACI with the STV prior. Then, an efficient iterative algorithm is employed for optimizing the proposed SCSTV model, where the ACI and the HRMS images are estimated and updated simultaneously. At last, reduced-scale and full-scale data fusion experiments visually and quantitatively verify the superiority of SCSTV.
Förderung
- National Natural Science Foundation of China
- China Postdoctoral Science Foundation
- Frontier Technologies Research and Development Program of Jiangsu
Zitieren
APA
Liu, P., Sun, L., Huang, N., Zheng, Z., & Xiao, L. (2026). Pansharpening With Structure and Coefficient-Driven Structure Tensor Total Variation Priors. IEEE Transactions on Geoscience and Remote Sensing, 64, 1-14. https://doi.org/10.1109/tgrs.2026.3669775
Vancouver
Liu P, Sun L, Huang N, Zheng Z, Xiao L. Pansharpening With Structure and Coefficient-Driven Structure Tensor Total Variation Priors. IEEE Transactions on Geoscience and Remote Sensing. 2026;64:1-14. doi:10.1109/tgrs.2026.3669775
BibTeX
@article{Liu2026,
author = {Liu, Pengfei and Sun, Liping and Huang, Nan and Zheng, Zhizhong and Xiao, Liang},
title = {Pansharpening With Structure and Coefficient-Driven Structure Tensor Total Variation Priors},
journal = {IEEE Transactions on Geoscience and Remote Sensing},
year = {2026},
volume = {64},
pages = {1--14},
doi = {10.1109/tgrs.2026.3669775}
}
EUVIMED