Supervised classification is a reliable method to classify the image into desired number of
classes of our interest. However, it requires knowledge of the remote sensing and the
spectral signature of various objects. Moreover, it is more effective if a person
involved has physically visited some locations in the image and can identify their type
clearly on the image. Many algorithms have been developed to assist the supervised
classification. We can choose from them depending upon our need, processing time and
accuracy level that we require.