Multi-sharpening hyperspectral remote sensing data by Multiplicative Joint-Criterion Linear-Quadratic Nonnegative Matrix Factorization

Abstract

Multi-sharpening consists in fusing a multispectral image with a hyperspectral one, to produce an unobservable image with the high spatial resolution of the former and the high spectral resolution of the latter. In this paper, a new fusion method, based on the spectral unmixing concept, is proposed. The proposed method, related to linear-quadratic spectral unmixing techniques, and based on linear-quadratic nonnegative matrix factorization, optimizes a new joint criterion by using new designed multiplicative update rules. This joint criterion exploits a spatial degradation model between the considered images. The proposed method is applied to synthetic data, and its effectiveness is evaluated with established performance criteria. Obtained results prove that the proposed method yields multi-sharpened hyperspectral data with good spectral and spatial fidelities. These results also illustrate that the proposed method outperforms the considered multi-sharpening literature approaches.