Change detectionn land use and land cover regional neyshabour using Different methods of statistical training theory

Change detection and identification Terrain is essential for understand of human interaction and the environment, that Aware of the cause is correct planning for sustainable development. Today, land use mapping manually is difficult with traditional methods and, remote sensing can help for Engineers in Land use mapping accurately and more quickly, and then assess the region changes. The purpose of this study to explore changes in land cover and land use by using Different methods of statistical training theory. In this study the process of preprocessing and data preparation to extract accurate information have been evaluated of three methods, Maximum likelihood and minimum distance and support vector machines by using the kappa coefficient. The results show that the maximum likelihood method with kappa 0/79, and overall accuracy of 29/83 than the minimum distance and support vector machine methods more accurately for made land use map .Then was produced land use map using by maximum likelihood method for the years 1988-200-2006 detection and evaluation the changes occurring by comparison post classification method. The most important changes to increased area of arable land, orchards increased of dam construction, urban land increased during the study period is 18 years.