Spectral bands prioritization for decision classes in remote sensing - Rough sets approach

In remote sensing, multi-spectral and multilayer imagery is providing good amount information about the targeted features. Using expert´s knowledge in the form of spectral signatures, one can optimize the information with respect to the requirement of the user. This paper deals with the mechanism of using expert´s knowledge in solving the dependency of bands on the terrain features by using Rough Set Theory. Alternatively, we are exploring from these digital numbers, the way an expert had given prioritization to different spectral bands.