عنوان مقاله [English]
The main aim of this study was to investigate forest cover changes in Ilam watershed from 1965 to 2006. The forest cover map was produced using a digital aerial photo mosaic, scale 1:20000 for 1965. The IRS-LISS III images were also employed to generate the map in 2006. After radiometric and geometric corrections, different image processing techniques, such as band rationing, tasseled cap transformation and principal components analysis were applied on the imagery to be entered in image classification process. The optimum bands set have been selected; using OIF (optimum indexing factor). Image classification was done using supervised rule (Maximum Likelihood and Minimum Distance) once for the 6 land use/cover classes i.e. forest, rangeland, agriculture, gardens, bare land and settlements and then for forest and non-forest classes. Accuracy assessment of forest-non forest mapping indicates that the highest overall accuracy is belonged to employ ML classifier (83% the overall and 78% the Kappa). Comparing the results in the period indicates that forest area in the region has been reduced about 16000 ha in a 42 years.