System for automated repair of city models

Overview

The use of 3D city models in simulations and spatial analyses has gained interest in recent years. However, practice has shown that geometric city models often contain errors that alter the simulation results. Troubleshooting is usually associated with a high manual effort. In order to reduce or even completely remove this effort, the project CityDoctor 2 was created.

Research Question

It is necessary to develop a coupled analysis and repair process with appropriate tools. Thus, a generally usable basic model in CityGML f-rmat with verified properties and a high degree of automation can be generated from an existing model, taking into account various application scenarios.

Procedure

Before a repair can take place, the errors and as much information about them as possible must first be identified and documented. With this information a repair can then be started. Two different approaches are planned. One approach is to develop complex deterministic algorithms that correct the errors. Furthermore, an evolutionary approach is pursued, which is based on a cyclical repair process. Together with an evaluation function, it comes as close as possible to a perfect repair.

Both error checking and repair should have a high degree of configurability to ensure control over the repair process.

ManagementProf. Dr. Margitta Pries (Beuth University of Applied Sciences Berlin), Prof. Dr.-Ing. Volker Coors
PartnerBeuth University of Applied Sciences Berlin, 3DIS GmbH, 3Dpartzz GmbH, Geoplex GIS GmbH, Mark Wewetzer, M.O.S.S. Computer Grafik Systeme Gmbh, con terra GmbH, Steinbeis Transfer Center Technical Consulting at the University of Applied Sciences Stuttgart, virtualcitySYSTEMS GmbH, State Office for Geoinformation and Rural Development Baden-Württemberg
Project Management AgencyVDI Technology Center
Websiteprojekt.beuth-hochschule.de/citydoctor2/
FundingFederal Ministry of Education and Research (BMBF)
Call for proposalFHProfUnt - Research at universities of applied sciences with companies
Duration01.11.2018–31.10.2021 extended until 31.12.21

Team

Name & Position E-Mail & Telephone
Prorektor Forschung und Digitalisierung+49 711 8926 2663 1/121
Akademischer Mitarbeiter