Characterizing Land use/land cover types by Landsat7 data based upon Object oriented approach in Kashan region

Remotely sensed data has high potential for characterizing land use/cover types. Traditionally, most of remote sensing image classification techniques are based on pixel-based procedures. In contrast to pixel-based procedure, image objects can carry more attributes than only spectral information. Object-based processing not only considers contextual information but also information about the shape of and the spatial relation between the image region. In this paper, we address the concepts of object-based image processing and presents an approach that integrates the concepts of object-based processing into the image classification and land use land cover type determination. The scheme proposed in this study is applied to classification of Landsat7 (ETM+) data of Kasha area. This study shows the applicapability of object-based approach for classification of Landsat7 (ETM+) data as well as show high overall accuracy (95%)of land use/land cover map. From the obtained results, we concluded that the main land cover types of the arid region could be discriminated with a high level of accuracy by object oriented approach