Mixed Noise-Oriented Hyperspectral and Multispectral Image Fusion

Hyperspectral images (HSIs) possess the capability to accurately characterize the attribute information of objects. However, they are usually obtained at a high spectral resolution with a compromise of their spatial resolution. In addition, they are easily contaminated by mixed noise induced by instrument and atmospheric effects. These disadvantages, to a certain degree, hinder the interpretations and applications of the HSIs. To overcome these limitations, in this article, we propose a novel mixed noise-oriented hyperspectral and multispectral image fusion method, termed (MixFus). First, a sparse noise detection method is proposed by leveraging a subset of specifically chosen hyperspectral bands to estimate noise in HSIs and then employing Gaussian mixture models (GMMs) to detect sparse noise from the estimated noise. Then, a robust subspace estimation method is introduced by replacing the detected sparse noise with new estimates using median values within a sliding window for a better estimation of the subspace, which offers improved accuracy and robustness of subspace estimation. Finally, in addition to the introduction of a state-of-the-art image prior based on the plug-and-play technique to exploit self-similarity characteristics in the eigen-images, we also impose a weighted group sparse regularization on the eigen-images to better promote the group sparsity of the spatial differences between the eigen-images, which further improve the denoising performance. We evaluate the proposed method by performing extensive experiments on three reduced-resolution HSIs and a full-resolution HSI in comparison with seven state-of-the-art competitors. Experimental results demonstrate the superiority of the proposed method over the competitors in the fusion of hyperspectral and multispectral images against mixed noise.