Enhanced Multilinear Mixing Model With Spectral Variability Awareness for Nonlinear Hyperspectral Unmixing
IEEE Geoscience and Remote Sensing Letters · 2026 · 23 · 5506205-5506205
DOI: 10.1109/lgrs.2026.3714033 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Geoscience and Remote Sensing Letters, 2026. Verfasst von Risheng Huang und 4 weiteren Autor:innen. Die Arbeit nennt 17 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Zhejiang Provincial Natural Science Foundation of China, Humanities and Social Sciences Youth Foundation, Ministry of Education.
Abstract (Originalsprache)
Hyperspectral unmixing (HU) aims to decompose mixed pixels into pure material spectra (endmembers) and their corresponding abundances. In complex real-world environments, spectral variability (SV) and nonlinear mixing commonly coexist, yet most existing HU methods address only one of these phenomena, limiting their performance in realistic scenarios. To overcome this limitation, we propose an enhanced multilinear mixing model with SV awareness (EMMSV) to jointly account for both effects. Building upon this model, we develop EMMSV-Net, a variational autoencoder (VAE)-based unmixing network employing three dedicated branches to learn scaling factors, perturbation terms, and scattering parameters, respectively. Experimental results on both synthetic and real datasets demonstrate that the proposed EMMSV-Net achieves robust unmixing performance compared to state-of-the-art methods, effectively handling the coupled effects of SV and nonlinear mixing.
Förderung
- National Natural Science Foundation of China
- Zhejiang Provincial Natural Science Foundation of China
- Humanities and Social Sciences Youth Foundation, Ministry of Education
- China Postdoctoral Science Foundation
Zitieren
APA
Huang, R., Zhou, H., Xiao, Q., Liu, Y., & Li, X. (2026). Enhanced Multilinear Mixing Model With Spectral Variability Awareness for Nonlinear Hyperspectral Unmixing. IEEE Geoscience and Remote Sensing Letters, 23, 5506205-5506205. https://doi.org/10.1109/lgrs.2026.3714033
Vancouver
Huang R, Zhou H, Xiao Q, Liu Y, Li X. Enhanced Multilinear Mixing Model With Spectral Variability Awareness for Nonlinear Hyperspectral Unmixing. IEEE Geoscience and Remote Sensing Letters. 2026;23:5506205-5506205. doi:10.1109/lgrs.2026.3714033
BibTeX
@article{Huang2026,
author = {Huang, Risheng and Zhou, Hao and Xiao, Qingjiang and Liu, Yufei and Li, Xiaorun},
title = {Enhanced Multilinear Mixing Model With Spectral Variability Awareness for Nonlinear Hyperspectral Unmixing},
journal = {IEEE Geoscience and Remote Sensing Letters},
year = {2026},
volume = {23},
pages = {5506205--5506205},
doi = {10.1109/lgrs.2026.3714033}
}
EUVIMED