International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 
  
lower than the original heights. Red coloured vectors refer to 
heights which became higher. The figure shows that most of the 
heights inside the lakes became higher. Most of the points 
which became lower are situated at the border of the lakes. 
Nevertheless, a big part of the differences of the left lake 
became lower, too. Here, one of the data sets seems to be coarse 
erroneous. The maximum differences between the original 
heights and the estimated heights are -1,84 m and +0,88 m, 
respectively. The right side of Figure 7 shows the result of the 
semantically correct integration with respect of the results 
without considering the semantics of the lakes (Figure 7, left). 
The semantically correct integrated data set shows that all 
constraints are fulfilled. The height values inside the lake and at 
the water line have the same level. The terrain outside the lake 
rises. Summarized, it could be stated that most of the residuals 
are rather small in respect of the vertical accuracy of the DTM 
of half a meter. The estimated lake heights are nearly identical 
to the mean values of the heights inside the lakes, the 
constraints are fulfilled exactly. > 
5. OUTLOOK 
This paper presents an approach for the semantically correct 
integration of a DTM and 2D topographic GIS data. The 
algorithm is based on a constrained Delaunay triangulation and 
a least squares adjustment taken into account inequality 
constraints. 
First investigations were carried out using simulated and real 
data sets. The objects used are lakes represented by a horizontal 
plane with increasing terrain outside the lake and roads which 
can be composed of several tilted planes. The results which are 
based on the use of different weights for the basic equations and 
equation constraints are satisfying. All predefined constraints 
can be fulfilled but a compromise between fulfilling these 
constraints and changing the terrain morphology has to be 
found. 
In the future the impact of blunders has to be investigated 
because height blunders or big differences to the equality and 
inequality constraints may cause a non-realistic change of the 
original height information of the DTM. 
Furthermore, the planimetric coordinates of the topographic 
objects were introduced as error-free. This may cause a 
erroneous height level of the topographic objects. Also the 
horizontal accuracy of the GIS objects has to be considered. 
REFERENCES 
Fritsch, D., 1985. Some Additional Information on the Capacity 
of the Linear Complementary Algorithm. In: E. Grafarend & F. 
Sanso, Eds., Optimization and Design of Geodetic Networks, 
Springer, Berlin, pp. 169-184. 
Fritsch, D., Pfannenstein, A., 1992. Conceptual Models for 
Efficient DTM Integration into GIS. Proceedings EGIS'92. 
Third European Conference and Exhibition on Geographical 
Information Systems. Munich, Germany, pp. 701-710. 
Heipke, C., 1986. Zum Vergleich von 2 Komplementaritits- 
algorithmen. Diploma thesis of the Technical University of 
Munich, Germany, unpublished. 
Klótzer, F., 1997. Integration von triangulierten digitalen 
Gelündemodellen und Landkarten. Diploma thesis of the 
Institute of Informatics, Rheinische  Friedrich- Wilhelms- 
Universität Bonn, Germany, unpublished. 
Koch, A., 2003. Semantically correct integration of a digital 
terrain model and a 2D topographic vector data set. 
Proceedings of ISPRS workshop on challenges in geospatial 
analysis, integration and visualization II, Stuttgart, 8-10 
September, CD. 
Lawson, C. L., Hanson, R. J., 1995. Solving Least Squares 
Problems. Society for Industrial and Applied Mathematics, 
Philadelphia. 
Lemke, C. E., 1968. On complementary pivot theory. In: G. B. 
Dantzig, A. F. Veinott, Eds., Mathematics in the Decision 
Sciences, Part 1, pp. 95-114. 
Lenk, U., 2001. - 2.5D-GIS und Geobasisdaten - Integration 
von Hóheninformation und Digitalen Situationsmodellen. Wiss. 
Arb. Fachr. Verm. Universität Hannover Nr. 244 and DGK bei 
der Bayer. Akad. D. Wiss. Reihe C, Nr. 546. PhD Thesis, 
Hannover, Germany. 
Pilouk, M., 1996. Integrated Modelling for 3D GIS. ITC 
Publication Series No. 40, PhD Thesis, Enschede, The 
Netherlands. 
Rousseaux, F., Bonin, O., 2003. Towards a coherent integration 
of 2D linear data into a DTM. Proceedings of the 21" 
International Cartographic Conference (ICC). pp. 1936-1942, 
Durban, South Africa. 
Schaffrin, B., 1981. Ausgleichung mit 
Ungleichungen. AVN, 6/1981, pp. 227-238. 
Bedingungs- 
Weibel, R., 1993. On the Integration of Digital Terrain and 
Surface Modeling into Geographic Information Systems. In: 
Proceedings 11” International Symposium on Computer 
Assisted Cartography (AUTOCARTO 11). Minneapolis, 
Minnesota, USA, pp. 257-266. 
ACKNOWLEDGEMENT 
This research was supported by the surveying authority of 
Lower Saxony Landesvermessung und Geobasisinformation 
Niedersachsen (LGN). We also express our gratitude to LGN 
for providing the data. 
  
Figure 7: Results of the integration process, left: non-semantic integration, right: semantically correct integration 
528 
KI 
Al 
Ot 
the 
tru 
int 
de 
pre 
int 
toc 
L. 
frc 
(ir 
tra 
ac