Analyze of the Relation between Cultural Areas and Urban Morphologies through remote sensing / Analyse des liens entre cultures et formes urbaines par utilisation de télédétection

Kurzfassung

From one country to the next, from a city to the next, the sensible feeling of the urban atmosphere differs. Shape, forms, shadows, and colors change. But the smaller the distance between two cities, the smaller the differences are. Nowadays, remote-sensing and earth observations products allow an analysis of cities at a global scale. The German Aerospace Center (DLR) recently produced two global datasets on urban morphology: an urban land cover classification and a set of boundaries for cities sprawl. This report presents the outcomes of my final internship at the Earth Observation Center of the DLR supervised by Dr. Hannes Taubenböck. This internship is the ending milestone of my engineer’s curriculum. My main goal during this internship was to develop and apply a framework to process those data in order to analyze the influence of culture on the morphological structure of cities. We led this survey on 110 cities all across the world. We propose here an approach based on an automated GIS preprocessing and an unsupervised clustering of the dataset. K-means and EM proved themselves to suite at best this task. The clustering they hence defined came close to the dividing of cultural spheres proposed by ethnologist and historian. Those results bring a new scope on the typologies of cities at both global and local scales.