Ability to prepare methods land use maps using satellite images (Case study: Kamyaran city)

It is important to have new land use plans in many areas, including natural resource management and land planning. Remote sensing data has a high potential for preparing up-to-date land use maps and land cover. The purpose of this study is to evaluate the methods of preparing land use maps of Kamyaran city using satellite images. In this study, OLI Landsat 8 satellite sensor data for June 2018 were used. Initially, preliminary processing, including radiometric, atmospheric, and geometric corrections, was performed on raw data. ground control points were used for training, accreditation, and land use mapping. The Landuse class was identified at each point by field survey and using Google Earth images in 9 user classes of agricultural lands, forest, garden, rich pastures, medium pastures, residential areas, water area, barren lands and rocky outcrops. In the following, maximum probability, minimum distance, support vector machine and Mahanalubi distance were used for the supervised classification in ENVI 5.3 software. To evaluate the accuracy of classification methods, two criteria of general accuracy and capa coefficient were used with ground control data. The results showed that the support vector machine method was 91.4% more accurate and the Kappa coefficient was 0.88% more accurate than other methods.