CI PA 2003 XIX 11 ' International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
from simple hand drawings, to CAD designs, GIS systems, 3D 
representations, animations and walkthroughs or even 
stereoscopic representations in virtual reality applications. 
Knyaz and Zheltov (Knyaz and Zheltov, 2001) discuss the 
different advantages of a virtual reconstruction. Three 
dimensional and virtual reality representations can be shared 
over the web or even be viewed through the web. Virtual 
museums are constantly increasing and new technologies are 
used for visualization such as photo realistic rendering. Instant 
sharing over the web with other researchers can help in 
exchanging opinions with people all over the world, working 
remotely. Easy accessibility to the public over the web or 
through educational and entertainment applications also 
promotes cultural heritage. Furthermore, searching 
manipulation and correlation of objects is simplified and the 
original artefacts or monuments are preserved. 
Figure 6. An overview of the ancient tomb used as case study 
The 3D representation has to provide adequate geometry 
characteristics and detailed enough textures for the 
archaeologists to be able to work on them. While the process of 
creating the virtual models can be complex, there are various 
techniques that try to automate the whole process as much as 
possible. 
4.2 The Visualization Process 
For the rendering of the model the stereoscopic display was 
used. It is a rear projection screen with mirror and 2 ordinary 
DLP's with polarizing lenses, Fig. 7. A 2.4 GHz PC with an 
Nvidia Quatro4 video card which has output for 2 monitors was 
used. The system renders one image for each eye, and both 
images are projected on the screen. The observer uses polarized 
glasses, which allows him to see only one image for each eye. 
The observers use polarized glasses, which allow them to see 
only one image for each eye, thus producing a stereo feeling 
especially when objects come out of the screen (negative 
parallax). For the rendering part the model was acquired from 
the photogrammetric methods and was textured using a texture 
map of the local stone and grass, in an attempt to show how the 
environment appears nowadays, Fig. 8. During the visualization 
process there are two options, to visualize the actual objects or 
to visualize the reconstructed objects. 
Figure 7. The visualisation set-up, showing the big screen and 
the mirror. The two monitors are used for control 
with each one giving a preview of what each eye 
will see on the big screen 
4.3 Animation 
The final model is loaded in the virtual reality system and the 
user can navigate around to examine the tomb. For navigation 
we use a Polhemus Fastrack tracker and an analogue joystick. 
Looking at the model in stereo and being able to move around 
gives a much more complete impression of the tomb and the 
size of the chambers. The virtual system can be further 
enhanced with collision detection to give some feedback about 
the geometry of the objects. A more advanced interface could 
allow the user to move objects around and try to fit different 
pieces together. 
Figure 8. A close-up of the tomb showing certain sections in 
different colour (with off-white texture). These are 
components that can be manipulated by the user, 
e.g. removed, to allow better visualisation 
Care should be taken though, as the correctness of the 
visualization can sometimes be taken for granted (Ogleby, 
1999). Different interpretations of archaeological data can result 
in different reconstructions and the virtual models should be 
questioned the same way the drawings on paper are questioned. 
In fact there is an on-going debate on whether striving for 
photorealism is the right approach when dealing with what is