CIPA 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
sherds with handles and sherds with a large curvature with 
respect to box 1 and box 2. Experiments have shown that 
the extraction of the profile line could be done on 5 out of 
26 sherds of box 3. Figure 11 shows a sherd containing a 
handle, where the estimated rotational axis is incorrect. 
Figure 11: Piece 8 from box 3 registered using an incorrect 
estimated rotational axis. The dark gray part is the outer 
view containing the handle. The light gray part is the inner 
view of the sherd. 
5 CONCLUSION AND OUTLOOK 
We have proposed an automated system for extraction of 
the profile line, which was required for archaeological clas 
sification and reconstruction. The work is part of a docu 
mentation system for ceramics. The acquired 3D-views 
have been registered and oriented using the rotational axis. 
Afterwards the profile line was extracted by intersecting 
planes from the registered and orientated 3D-views of the 
sherd. The method has been tested on real data with rea 
sonable good results. Also sherds that can not be processed 
manually, because of their low curvature, can be processed 
by the presented system. 
Future work will go towards making the existing system 
more robust with respect to the sherds with handles and 
bottom fragments, so that the registration can be applied 
on a larger percentage of sherds. The presented system 
is able to determine, when the estimation of the rotational 
axis fails, but it can not detected the reason (e.g. handle) 
for that. So a detection and separate processing of sherd 
with handles or bottom-pieces would increase the perfor 
mance of the system. 
As the rotational axis leads to the position of the fragment 
in the unbroken vessel multiple fragments will be matched 
to reconstruct the whole object. Figure 12 shows a whole 
object, which was reconstructed manually by archaeolo 
gists. This could be done by matching the profile lines 
from different sherds of one object, so that the whole ob 
ject can be reconstructed. 
ACKNOWLEDGEMENTS 
This work was partly supported by the Austrian Science Founda 
tion (FWF) under grant P13385-1NF, the European Union under 
grant 1ST-1999-20273 , the Austrian Federal Ministry of Edu 
cation, Science and Culture and by the innovative project ’3D 
technology’ of the Vienna University of Technology. 
Figure 12: Complete manually reconstructed vessel. 
REFERENCES 
Adler, K., Kampel, M., Kastler, R., Penz, M., Sablatnig, R. and 
Hlavackova-Schindler, K., 2001. Computer aided classification 
of ceramics - achievements and problems. In: W. Borner and 
L. Dollhofer (eds), Proc. of 6th Inti. Workshop on Archaeology 
and Computers, Vienna, Austria, pp. 3-12. 
Bennett, J. and MacDonald, J., 1975. On the Measurement of 
Curvature in a Quantized Environment. IEEE Trans, on Com 
puter 24, pp. 803-820. 
Cosmas, J., Itagaki, T., Green, D., Grabczewski, E., Gool, L. V., 
Zalesny, A., Vanrintel, D., Leberl, F., Grabner, M., Schindler, K., 
Karner, K., Gervautz, M., Hynst, S., Waelkens, M., Pollefeys, M., 
DeGeest, R., Sablatnig, R. and Kampel, M., 2001.3D MURALE: 
A multimedia system for archaeology. In: Proc. of Inti. EuroCon- 
ference on Virtual Reality, Archaeology and Cultural Heritage, 
Athens, Greece, pp. 297-305. 
Degeest, R., 2000. The Common Wares of Sagalassos. In: 
M. Walken (ed.), Studies in Eastern Mediterranean Archaeology, 
III. 
Kampel, M. and Sablatnig, R., 1999. On 3D modelling of archae 
ological sherds. In: Proceedings of International Workshop on 
Synthetic-Natural Hybrid Coding and Three Dimensional Imag 
ing, Santorini, Greece, pp. 95-98. 
Leute, U., 1987. Archaeometry: An Introduction to Physical 
Methods in Archaeology and the History of Art. John Wiley & 
Sons. 
Mara, H., Kampel, M. and Sablatnig, R., 2002. Preprocessing 
of 3D-Data for Classification of Archaelogical Fragments in an 
Automated System. In: F. Leberl and F. Frauendorfer (eds), 26th 
Workshop of the Austrian Association for Pattern Recognition 
(OeAGM/AAPR), Graz, Austria, pp. 257-264. 
Matas, J., Shao, Z. and Kittler, J., 1995. Estimation of Curva 
ture and Tangent Direction by Median Filtered Differencing. In: 
8th International Conference on Image Analysis and Processing, 
pp. 83-88. 
Nadeau, D., 1999. Building virtual worlds with vrml. In: Com 
puter Graphics and Applications, IEEE, pp. 18-29. 
Orton, C., Tyers, P. and Vince, A., 1993. Pottery in Archaeology. 
Cambride University Press. 
Poblome, J., 1999. Sagalassos Red Slip Ware. In: M. Walkens 
(ed.), Studies in Eastern Mediterranean Archaeology, II, Brepols. 
Sablatnig, R. and Kampel, M., 2002. Model-based registration of 
front- and backviews. Computer Vision and Image Understand 
ing 87(1-3), pp. 90-103.