Advantages of Nonlinear Intensity Components for Contrast-Based Multispectral Pansharpening

In this study, we investigate whether a nonlinear intensity component can be beneficial for multispectral (MS) pansharpening based on component-substitution (CS). classical CS methods, the is linear combination of spectral components and lies hyperplane in vector space that contains MS pixel values. Starting from hyperspherical color (HCS) fusion technique, devise novel method, which hyper-ellipsoidal surface instead surface. The proposed method insensitive to format data, either floating-point radiance values or fixed-point packed digital numbers (DNs), thanks use multivariate regression between squares interpolated bands squared lowpass filtered Pan. MS, Euclidean radius used by HCS, makes no longer lie hypersphere samples, but hyperellipsoid. Furthermore, before accomplished, are corrected atmospheric haze, order build multiplicative injection model with approximately de-hazed components. Experiments GeoEye-1 WorldView-3 images show consistent advantages over baseline HCS performance slightly superior those some most advanced methods.