Hyperspectral and panchromatic image fusion based on CNMF

Abstract

Hyperspectral images contain rich spectral features, but the spatial resolution of hyperspectral images is low, so image fusion technology in the same scene plays an essential role in satellite imaging. The commonly used method now is to transfer the spectral information in the hyperspectral image to the panchromatic image, but many algorithms cannot avoid spectral distortion. Based on the coupled non-negative matrix decomposition (CNMF) algorithm, a spectral constraint regularization term is introduced to avoid spectral distortion and maintain spectral integrity. The experimental results were compared with the other four most advanced methods, this method has obvious advantages in terms of visual effects and evaluation indicators.