In: Stilla U, Rottensteiner F, Paparoditis N (Eds) CMRT09. IAPRS, Vol. XXXVIII, Part 3/W4 — Paris, France, 3-4 September, 2009
115
GRAPH-BASED URBAN OBJECT MODEL PROCESSING
Kerstin Falkowski and Jürgen Ebert
Institute for Software Technology
University of Koblenz-Landau
Universitätstr. 1, 56070 Koblenz, Germany
{falke|ebert} @uni-koblenz.de
http://www.uni-koblenz-landau.de/koblenz/fb4/institute/IST/AGEbert
KEY WORDS: Urban, Model, Metadata, Data Structures, Algorithms, Processing, Services.
ABSTRACT:
Urban object models are valuable assets that allow reuse in different applications. Besides the need for exchange formats there is
also the need for comprehensive, efficiently processable data structures for such models. This paper presents a graph-based schema for
integrated models of urban data, that is an adaption of the comprehensive CityGML approach. It defines an explicit graph representation
and thus is well-suited to efficient processing algorithms. The paper demonstrates how appropriate light-weight components realizing
different kinds of services on models can be used for consistently processing semantics, geometry, topology and/or appearance of
graph-based models compliant to that schema. Several examples are given.
1 INTRODUCTION
Urban models are valuable assets that should be constructed once
while being used multiple times in different applications. There
fore the exchange of 3d city models between different tools is
indispensable. Various XML formats are being used to achieve
interoperability between tools. These formats (e.g. CityGML
(Groeger et al., 2008)) are able to carry topological, geometric,
semantic, and appearance information, but in different forms and
to varying extent.
Applications, like tools for the automatic extraction of topo
graphic objects, build on these urban object models and improve,
transform, and analyze them in different ways. XML-technology
(e.g. XSLT and XQuery) is widely used to support these activ
ities, but this technology is not well-suited for the implementa
tion of the various algorithms on urban objects which come from
the areas of algorithmic geometry, computer graphics, and im
age recognition, since the necessities of efficient content-based
traversal of all relevant information is only hard to realize in the
essentially tree-like structures supplied by XML.
Therefore, a comprehensive, efficiently processable data struc
ture for urban objects is essential. Geographic information sys
tems share this necessity with route guidance systems, where a
graph-like internal representation of data is used for the compu
tation of routing information.
In this paper, we present an approach for the efficient storage,
analysis, and manipulation of city models using graphs and for
the development of application specific components collectively
working on an integrated, efficient graph representation of city
models. Import/export from/to CityGML is tackled, as well. 1
After a short overview of the state of the art in section 1.1, sec
tion 2 shortly introduces the employed graph and component
concepts. Section 3 describes the graph-based integrated model
schema with all its aspects, and section 4 shows how quite differ
ent kinds of functionalities can be implemented on such a model
by independent components. Section 5 concludes the paper.
x The project is funded by the DFG (EB 119/3-1).
1.1 State of the art
There are several XML-based modeling languages for urban ob
jects. The City Geography Markup Language (CityGML) 1 is a
common information model for the representation of 3d urban
objects and an official standard of the Open Geospatial Consor
tium (OGC) since August 2008 (Groeger et al., 2008). Besides
representing geometry, CityGML can also be used to model topo
logical and semantic properties of 3d city models and to attach
appearance information like textures.
Models described using CityGML can be rendered by Ifc-
Explorer for CityGML' from the Institute for Applied Com
puter Science, Forschungszentrum Karlsruhe or the LandXplorer
CityGML Viewer 2 3 4 5 6 from Autodesk and by Aristoteles ? from the In
stitute for Cartography and Geoinformation, University of Bonn.
Besides CityGML there are other languages for the representation
of 3d urban objects. One common approach is the OGC standard
Keyhole Markup Language (KMLf. CityGML uses a subset of
the OGC standard Geography Markup Language (GML) (Cox
et al., 2001) for geometry representation, KML derived his ge
ometric elements from GML. KML is often combined with the
COLLADA 7 8 exchange format for 3d assets. Another 3d model
ing language is Extensible 3D (X3Df, the successor of the Virtual
Reality Modeling Language (VRML) standard.
2 BASIC TECHNOLOGIES
2.1 TGraph technology
For the efficient manipulation of urban object models with all
their aspects a versatile and powerful basic technology is needed.
In the context of this work TGraph technology is used.
2 http://www.citygml.org, http://www.citygmlwiki.org
3 http://www.iai.fzk.de/www-extern/index.php?id=1570
4 http://www.3dgeo.de/citygml.aspx
5 http://www.ikg.uni-bonn.de/aristoteles
6 http://www.opengeospatial.org/standards/kml
7 http://www.khronos.org/collada
8 http://www.web3d.org/x3d
