Comparing fuzzy and maximum likelihood methods to land cover mapping in Gandoman wetland using Landsat satellite data

Preparation of land use and land cover maps in order to inform from land use and planning with optimal planning one of the basic measures to achieve sustainabll developement. This issue is more important in sensitive areas and particularly in wetland due to dependence of hydrologic and biogeological performance of them to surface and groundwater flows and their status in the landscape. Nowdays, many algorithms have developed to compare maps of land use / cover maps. The aim of this study is comparison of two methods of maximum likelihood and fuzzy method for preparation of land use/cover maps in Gandoman international wetland and the land around it using Landsat images. For this study, Images (2014) were used related to landsat. Based on obtained results the study area was classified into farmland, bare land, rangeland and lagoon. Thereafter overall accuracy and kappa computed for each land use. Finally results showed that fuzzy method more accurately produced land were used than maximum probability method. The result can be staded maps of land use fuzzy algorithm with overall accuracy 86،70 and kappa coefficient 0.79 to maximum likelihood in overall accuracy 81.20 and kappa coefficient is 0.71, higher accuracy.