The primary objective of spectral signature analysis was to check the seperability or to evaluate the spectral separation between landcover classes. The reflectance curves for all bands were generated to study the spectral response pattern of each landcover classes. Line graphs were generated to show the spectral seperability of different landcover classes. In an optical image, vegetation can be easily distinguished from it’s unique spectral signature. The reflectance of the tea patches and other vegetation classes showed high peak in the NIR band. The low peaks shown at the Red region of the spectrum are mainly due to the absorption of chlorophyll. Very importantly it was found that river bed is showing its peak in the Red, NIR and Green region of the spectrum. This may be due to higher reflectance and lower absorption.  Figure 14: Graph Showing The Landcover Class Seperability in LANDSAT ETM+  Figure 15: Graph Showing The Landcover Class Seperability in LISS III  Figure 16: Graph Showing The Landcover Class Seperability in ASTER The three spectral curves, showed that ASTER has highest peak for healthy, moderately affected and diseased patches in the (NIR) region. Lowest reflectance was obtained from the LISS III image. |