Representative sample collection is the most time consuming process in the classification effort. In this study, samples were randomly selected from known areas using the ‘region of interest’ (ROI)tools provided by ENVI version 4.8 (ITT industries Inc., Boulder, CO,USA) with the help of ground survey and Google Earth tool. Finally,70 and 138 ROI regions were selected for forest and non-forest class samples, respectively. The total sample pixels for forest and non-forest class were 13,212 and 15,824 pixels. The distribution of the sample pixels was uniform, which made it well representative for the whole study area. Half of the sample pixels were randomly selected as training samples, and the remaining half as validating samples. The training and validating samples had no overlap.