On Constructing a Base Map for Collaborative Map Generation and its Application in Urban Mobility Planning

Maik Drozdzynski, Stefan Edelkamp, Andreas Gaubatz, Shahid Jabbar, Miguel Liebe

Abstract: Urban mobility planning depends largely on the presence of good navigation data in the form of vectorized maps. Unfortunately, vector maps are not always available for many areas – especially for many of the third world countries. On the other hand, good paper maps collected by the city authorities are widely available. A solution is the <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">collaborative map generation</i> process that allows people to share the collected GPS traces. Nevertheless, the integration of these GPS traces is itself a challenge and requires a good <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">base map</i> . This paper presents a method to extract calibrated road topology from raster maps to provide such a <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">base map </i> for collaborative map generation process. Our approach takes a bitmap and uses different graphics filters to infer the road geometry. We propose an aggregation algorithm that extracts the actual vectorized road fragments and construct a graph of road network. To evaluate the proposed algorithms, we have implemented them and tested on real raster maps collected from the city authorities of Dortmund, Germany. We also report the integration of our approach into SUMO, a state-of-the-art traffic simulation tool for urban mobility.