Analysis of pansharpening methods for multispectral satellite imagery

The article examines the approaches of pansharpening, which is one of the effective methods of creating large-scale high-resolution and large-scale multispectral images based on combining small-scale multispectral and PAN images. The methods were analyzed by dividing into CS, MRA, DM and DNN classes according to calculations. The methods used in the classes are analyzed, and their mathematical basis, computation plan, and perspectives on how to use them are explained. Research problems have been created in an effort to remove issues with practical problem-solving techniques. The study specifically addresses the problems of creating hybrid approaches based on CS and MRA methodologies, as well as the identification of activation and error functions that produce useful outcomes in DNN pansharpening methods.