The Application of Object- Based Image Analysis Approach for Land Use/Cover Change Modeling of Urban Growth within the City of Maragheh

Land-use can be considered as a combination of physical, social, cultural, economic, and informational concept of every country. In fact, the land-use maps include ways for using land for human’s different needs. Since warning land-use maps and their updates are as the most important goal in the management of lands, the use of remote sensing technology is the best way to extract the user's maps. The present study aims to employ HDR sensor with digital processing SPOT satellite images related to 2005 and satellite image obtained from the sensor of AVNIR, AlOS for 2011 to extract land use maps based on object-based image analysis approach. Accordingly, in the pre-processing step was spent on the image and in the processing phase after the detection of images, they were classified using object based advanced method. For the classification the function (the nearest neighborhood) and algorithms (hierarchical classification) were used. The method of image processing was object-based which was performed using spatial and spectral algorithms in knowledge-based manner at e-cognition software. In the next step, the validation step was performed and the accuracy of classification for 2005 was calculated to be 84 percent and for the year 2011 it was assessed to be about 81 percent. The results of this research are important for decision makers in this area for the task of regional planning and monitoring.