Deutsch Intern
Chair of Computer Science I - Algorithms and Complexity

Map Matching and Map Conflation

Geographic Information Systems allow users to combine data from different sources. For example, by applying map matching algorithms, GPS trajectories can be combined with road network data to infer information on road names and adjoining links. Due to geometric errors of the trajectory and the road network data, map matching is difficult. Moreover, GPS may be unavailable, for example, due to occlusions in tunnels. A promising approach is therefore to equip cars with additional sensors, for example, cameras or laser scanners, which allow objects to be detected and to be matched with features in a geographic information system. Such objects can be, for example, poles of traffic signs or traffic lights.

We also develop algorithms for matching geospatial data sets of different scales, which due to map generalization may contain very different contents. For example, while in large-scale maps rivers are represented as area features, in small-scale maps they are represented as line features.


Vehicle Localization by Matching Triangulated Point Patterns.
In: O. Wolfson, D. Agrawal und C.-T. Lu, editors, Proceedings of the 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM-GIS'09), November 4-6, 2009, Seattle, WA, USA, pages 344-351. 2009.
J.-H. Haunert and C. Brenner.
[doi] [PDF] [BibTeX] 

Matching River Datasets of Different Scales.
In: M. Sester, L. Bernard and V. Paelke, editors, Advances in GIScience: Proceedings of the 12th AGILE International Conference on Geographic Information Science, June 2-5, 2009, Hannover, Germany, series Lecture Notes in Geoinformation and Cartography, pages 56-63. Springer, Berlin, Germany, 2009.
B. Kieler, W. Huang, J.-H. Haunert and J. Jiang.
[doi] [PDF] [BibTeX]