- 141 - 
characteristics of the objects can be used for further analogical 
or numerical computer assisted interpretation and automatic 
features extraction. But no such device is already available now. 
The other principle that allows optical or optoelectronic 3D 
recording of archaeological objects is the triangulation, i. e. the 
fact that the distance between a point and an observer can be 
determined through the change of position caused by this point 
to a laterally projected light pencil. This principle is used in 
many relief analysis techniques, among which the very well- 
known linear scanning method. 
The linear scanning method can be summarised as follows: a 
laser line is projected on the studied object and appears distorted 
by the relief of this object when it is seen with an angle from the 
incident direction. By scanning the object surface with an 
accurate positioning system, it is possible to get the depth 
information line by line and then to reconstruct a 3D 
representation of the object. This technique may only examine 
one scan at a time. So it is limited by the accuracy of the 
positioning system and the way to create the scanning line. 
Many tests have revealed that, because of this limitation, it is 
very difficult (and sometimes impossible) to use a laser 
scanning system with high precision in aggressive 
environmental conditions like those found in many 
archaeological sites. The possible applications in the field of 
Archaeology are thus limited. 
The main problem of the scanning method, that is the problem 
of moving precisely the scanning laser line, can be eliminated 
by generalising the triangulation principle: instead of one single 
line, it is a constructed pattern or a grid that is projected. Its 
deformation is then used as the probe of the relief. This light 
pattern has to be periodically structured and static, based on a 
grid alternatively light and dark. One projects it on the surface 
to analyse. By recording the scene with a CCD camera, it is 
possible to superpose the image of the grating modified by the 
relief with the reference one without any deformation. This 
process, which replace the comparison between the distorted 
laser line and the original one, creates geometrical shapes 
(Figure 2); it is the moiré effect, whose fringes pattern is closely 
connected with the relief of the analysed object. Everyone has 
already seen this effect, often ignoring it: when somebody on 
television wears a striped or squared clothe, this pattern, 
modified by the anatomy of the person itself, interferes with the 
pixels grid of the camera and generates a moiré effect. The 
interpretation of these moiré pictures through the triangulation 
principle gives the whole relief information of the analysed 
object (Figure 3). The accuracy of this technique is comparable 
with the accuracy of the traditional laser scanning but its 
process is much more faster, since a surface of one square meter 
can be analysed in one time, with a single shot. Besides, the 
elimination of the problems linked to the precise and regular 
moving of the laser line makes it much easier to use in difficult 
in situ conditions. So the moiré technique perfectly fulfil the 
requirements of archaeological recording: fast acquisition, 
accuracy, robustness and flexibility, necessary to allow working 
on site, in aggressive environmental conditions. Moreover, by 
combining moiré and photogrammetric approaches, it is 
possible to define the recorded 2.5 surface shape in an euclidian 
cartesian reference system that allows metrology and merging 
thematic and geometric information from the object in this 
system. 
Figure 2. Moiré effect on an Ancient Egyptian Relief 
(copy of the relief Brussels MRAH E 2157) 
Figure 3. Detail of a 3D recording of the same relief with the 
projected moiré technique 
5. CONCLUSION: THE AIMS OF THE OSIRIS 
PROJECT 
In conclusion of all this, the ICAUL (Interdisciplinary Center 
for Archaeometry of the University of Liège), with Hololab 
Laboratory and SURFACES (Service Universitaire de 
Recherches Fondamentales et Appliquées en Cartographie et 
Étude Spatiales) Laboratory, of the University of Liège, have 
decided to develop together a complete portable set-up (making 
the whole optoelectronic acquisition and the data processing) 
specifically dedicated to the quick and accurate numerical 3D 
recording of archaeological documents. It will use the projected