Land use/land cover classification for applied urban planning - the challenge of automation

The number of remote sensing platforms increased during the last decades significantly and produced an ascend of geographical data availability at variable scale. The development of last generation satellites leads to a revival in sensor research. The highest spatial resolution commercial satellite systems even challenges the aerial photogrammetry. Increasing knowledge of methods for analyzing and processing remote sensing data is not equipollent to the outcomes of new remotely sensed geospatial data. Especially very high resolution (VHR) remote sensing data gains the application range for urban remote sensing (URS) and becomes therefore interesting for urban planners. Due to rapid urbanization processes and growth of population, urban sprawl becomes a challenging task for urban planners. Besides land cover, land use is one of the most important information for planning authorities and their decision-making procedure. A classification strategy for applied urban planning was developed to gain information about urban environment. In this context automation and transferability of algorithms are crucial factors for user oriented solutions. Therefore a problem-solving approach with object-based classification algorithms was elaborated. The developed classification procedure lead to land cover and land use maps in a semi-automated approach.