AN ANISOTROPIC VARIATIONAL PANSHARPENING MODEL WITH ADAPTIVE COEFFICIENTS

Pansharpening is a widely used technique in the field of remote sensing, which aims to obtain a fused product with both high spatial and spectral resolution. In this paper, we propose an adaptive-coefficients-based anisotropic variational model for pansharpening. First, a panchromatic (PAN)-guided anisotropic regularization term is proposed, which can integrate the spatial information of the PAN image into the fused product. Then, to better characterize the relationship between the PAN image and the fused product for reducing spectral distortion, we propose a PAN constraint term with adaptive mixing coefficients. Combining these two terms with a conventional spectral fidelity term, a new variational panchromatic sharpening model is formulated. We prove the existence and uniqueness of the minimizer for the proposed variational model and design an efficient finite difference scheme with optimized rotation invariance to solve this model numerically. To demonstrate the effectiveness and robustness of the proposed pansharpening method, we conduct extensive experiments on reduced-resolution and full-resolution datasets. Experimental results indicate that the proposed method is superior to other state-of-the-art approaches in terms of both quantitative and qualitative assessments.