Analysis of Most Significant Bands and Band Ratios for Discrimination of Hydrothermal Alteration Minerals

Abstract

Band ratios are one of the most common features for discrimination and identification of minerals using remotely sensed data. The band ratios to be used are determined mainly by spectral analysis and trial experiments. In the present work, we systematically examine the best features and best band ratios using Bhattacharyya distance based optimal band analysis. The experiments indicate that the ASTER nine bands and band ratios within the wavelength range of 2200 nm to 2300 nm are significant for hydrothermal alteration minerals