Resumen
In Germany, a dataset of 3D building models has been created at two levels of detail, based on the nationwide standards of the Working Committee of the Surveying Authorities of the States of the Federal Republic of Germany (AdV). This work has been done by the German federal states. There are some interesting details in the datasets; for example, in Bavaria, the building ground plans are taken from the official cadastral map. Both the 3D building model of Bavaria and the Infrastructure for Spatial Information in the European Community (INSPIRE) directives for the data specification of buildings (INSPIRE-BU) are based on the Open Geospatial Consortium (OGC) CityGML standard. Thus, an obvious development was to investigate the possibility of converting these data to a schema conforming to INSPIRE. In order to convert these 3D building models at the second level of detail to INSPIRE, a project group of the AdV entitled ?3D-Geobasisdaten? created an initial mapping table, based on the AdV-CityGML profile and GeoInfoDok 7.0 beta. To ensure that the schematic transformation is operational, 100% compliance between the source data and the source schema is required. The 3D-Geobasisdaten group and the Central Office for House Coordinates, Building Polygons and 3D Building Models (ZSHH) have jointly developed relevant criteria for data at both levels of detail. Tests on 3D building models at both levels of detail were carried out to check the conformity of the schema and profile, as well as the geometric and semantic data checks defined by the AdV. During the operational work at the Bavarian Agency for Digitization, High-Speed Internet and Surveying, it was shown that errors can reliably be detected. In an extension of this work, a joint research project between the former Bavarian State Agency for Surveying and Geoinformation and the Chair of Geoinformatics at the Technical University of Munich (TUM) provided proof of the convertibility of 3D building models based on AdV standards to INSPIRE specifications via conversion of a sample dataset.