Model-based view at multi-resolution image fusion methods and quality assessment measures

摘要: ABSTRACTWe propose to look at multi-resolution image fusion or pan-sharpening task from a model-based perspective. Explicit definition of all models assumptions used in the derivation method allows us understand rationale properties existing methods and shows way for proper usage proposal/selection new better satisfying needs particular application. Earlier mentioned property ‘better’ should be measurable quantitatively, e.g. by means so-called quality measures. The difficulty assessment is that reference missing. Existing measures protocols are still not satisfactory because quite often valid fulfilled. From view, it follows naturally measure can defined as combination error model residuals using common general assumed methods. It shown most metho...