4 . TEXTURE MAPPING USING CALIBRATION RESULTS 
The cloud of points can be textured by using the simultaneously taken images with the help of calibration parameters. First, the cloud 
of points had to be triangulated to get a boundary representation of the object. Then the vertices of each triangle were mapped to each 
image by using Eqn. 5 and 6. This equation realizes the relationship between the object and the image (pixel) co-ordinates. The three 
pixel co-ordinates of a triangle cuts out an area of the image which is then used to texture the corresponding triangle in the object. 
Some triangles refer to two or more images because they overlap one another. Unfortunately, corresponding areas of adjacent images 
are often taken with very different exposures (Fig. 3) an intelligent mean value operation is thus necessary to find the image with the 
correct exposure. The Fig. 8c is a first result from the application of supplementing scanner and photogrammetric data. 
points by ray tracing c) Automatic generated 3d model of cloud of points with texture 
(POV-Ray Team 1999) mapping (clipping with 9 from total 22 images Fig. 3 
Figure 8: Texture matching with using the 3D information from point cloud 
With the availability of a complete 3D geometric model of the object it is possible to reconstruct the circumstances of image 
exposure during scanning process. Fig. 8a shows the original video image and in Fig. 8b, a visualization of the cloud of points made 
by using a ray-tracing program. The shadows in both images look similar. Why not use this information to correct the original 
image? Textures which are independent of the actual light exposure are very valuable for computer visualization and animation 
because the light sources are simulated in virtual reality processes too. 
The institute for geodesy and photogrammetry, technical university of Braunschweig is working on a project for virtual 
reconstruction of the Cistercian abbey sited in Walkenried / Harz. Up to now the first phase of the project is complete, but not all 
existing buildings have been reconstructed. It is planned to use the supplementing laserscanning method in the second project phase 
to complement the computer animation. 
5. CONCLUSIONS 
A calibration technique for the Callidus system, which allows supplementing geometric data from laser scanner with images, has 
been explained. The combination of photogrammetric and laser scanner systems compliments one another perfectly for a complete 
object modelling. Today only the Callidius system can capture images during the scanning process however only with poor quality 
and resolution. In future video cameras have to be replaced with digital cameras with minimum resolution of 2 mill, pixel in order to 
attain more accurate and better quality results that can be used for different purposes. The use of this technique in cultural heritage 
documentation has been presented. 
6. ACKNOWLEDGE 
I would like to thank Dirk Mittelstadt and Olaf Gay, ÖbVI Mittelstadt & Schröder, for support, data and fruitful discussions. 
REFERENCES 
Fellner, W., 1992. Computergrafik. Number 58 in Reihe Informatik. Bl-Wissenschaftsverlag, Mannheim, Leipzig, Wien, Zürich, 
second edition. 
Kern, F., 2001. Xdesy. www-public.tu-bs.de:8080/7fkern/. 
Luhmann, T., 2000. Nahbereichsphotogrammetrie. Wichmann, Heidelberg.