Urban Boundary Mapping Using Sentinel-1A SAR Data

Mapping urban boundaries is an area of considerable research interest. The ability to monitor growth and change is facilitated by the ease and accuracy of boundary delineation. This research proposes an open-source (data and software) workflow for detecting and delineating the urban-rural boundary. Using Sentinel-la SAR data for the city of Melbourne, Australia and a combination of texture and k-means classifications, using ESA SNAP software, a three-category map is produced that distinguishes urban from rural areas. A simple raster to polygon conversion produces a useable shapefile that can be added to a broader GIS based analysis.