Spectral separability of vegetation classes in a dry tropical region of India using IRS-1A LISS-1 data

Abstract

This paper describes the spectral separability of IRS-1A LISS-1 data for vegetation classification. Separability was assessed for enhanced data (principal component analysis and hue saturation intensity transformation) as well as for raw IRS bands. The analysis indicated that the enhanced data do not show a superior performance in comparison to the 4 bands of IRS data. The comparison of bands 2, 3 and 4 of IRS with the enhanced data showed that the three PCA bands as well as the HSI transformation bands increase the vegetation class separability significantly.