The application on Image Processing of Remote Sensing and Geographic Information Systems: Change detection of Eskişehir city related to rail systems

Cities have become uninhabitable places due to climate change caused by environmental pollution, which is the effect of greenhouse gases. Irregular migration to urban centers and unplanned construction resulting from this migration, made urbanization unsustainable. In order to make urbanization planned and sustainable, it is necessary to rearrange the structuring by using technological data. One of the modern technologies required for urbanism is the discipline of Remote Sensing and Geographic Information Systems. Transport networks should be redesigned by using this discipline. Nowadays, there is a trend to use rail systems as public transportation in the modern world with environmental sensitivity. Public transportation by using Remote Sensing and Geographical Information Systems by rail systems is important for sustainability of urbanization. In this thesis, in order to determine the urbanization change of Eskişehir province according to rail systems between 2006-2019, change detection is performed by machine learning models. The rail influence region of Eskişehir has been extracted by using ArcGIS software. In the rail influence region, segmentation performance on two different classifiers was compared with the proposed descriptor. In this way, change detection was carried out by marking the structure and rural areas using the high performance classifier. Key Words: Change Detection, Geographic Information Systems (GIS), Remote Sensing (RS), Fisher Linear Discriminant Analysis (FLDA)