City Models - Automation in Research and Practice

This paper concentrates on the role of automated extraction of city models in today’s practical work. The state of research and practical implementations is compared and the question is raised why the degree of automation in modelling systems is still very low – despite many years of scientific research efforts. Some of the problems are identified, in particular the need to achieve a high degree of automation and to assess the advantage of automatic components by looking at the effects on the entire processing chain rather than only single stages. Three examples for automation are given. The first two are approaches for the reconstruction of building geometry which integrate 2D ground plan information with digital surface models. A heuristic subdivision of ground plans into rectangles as well as a rule-based reconstruction relying on discrete relaxation and constrained tree search are discussed. The third example shows how quality control can be automated using digital surface models.