First, an unsupervised classification was performed on the SSC Image using the ISODATA clustering method to classify the image into the desired classes (30) in the final output.
A thematic raster layer was generated using the ISODATA algorithm while running ERDAS IMAGINE
Thirty classes were derived in the classification with:
maximum number of iterations set to 12
convergence threshold set to 0.95
The pixels were identified for each of the categories and they were grouped into land cover categories:
water, shadow, coniferous trees,deciduous trees, mixed trees, scrub grass, grass, and urban