CIP A 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
507 
ermany, 
e Artifacts. 
n top soils of 
rchaeological 
;arch project, 
ml collection, 
lete 3D point 
irocess which 
ns have to be 
tee points are 
ight, place of 
GIS. Certain 
This policy 
comparisons. 
I data storage 
sols for local 
'or inspection 
irchaeologists 
ine all these 
eolithic Land 
; aim of this 
; cases, huge 
localities has 
Terent collec- 
nowing from 
agether again 
lole material 
problem that 
how to make 
ists and other 
Because the 
erial, with an 
tional way of 
t to develop a 
fic evaluation 
s. This was a 
n. The docu- 
ing. The scan 
e linked to a 
>) which may 
;ked up and 
he methods, 
be used for 
2. OBJECTS AND MATERIALS 
The stone artifacts from Rheinhessen which were studied up to 
now were all manufactured from local raw materials like quartz, 
different kinds of Devonian quartzite and Tertiarien quartzite, 
by hard hammer percussion. That means in general, one rock is 
hit by another. Normally, a well rounded cobble is used as a 
hammer stone. The rock from which the flakes are removed is 
called core. The striking products are the flakes. The best angle 
for striking stones is under 90°. All flakes have two sides. The 
dorsal surface represents the outside of the core when the flake 
was made (which may show cortex or negatives from other 
flakes removed before as well). The other side of a flake is 
called the ventral surface. This is the old inner side of the core. 
Just below the striking platform, which is the surface where the 
hammer stone struck the core and knocked off the flake, you 
find on the ventral side the bulb of percussion which represents 
the energy waves of the flaking blow as they spread into the 
core. On the whole ventral surface you find wallner waves 
(ripple marks) which radiate in series of progressively larger 
arcs from the point of percussion up to the opposite (distal) end 
of the flake. These waves show the striking direction. On the 
cores you find the negatives of the flakes which have been 
removed. The ripple marks in these negatives also allow an 
analysis of the striking direction and the technique which was 
used. A normal byproduct from striking stones are small (less 
than 1,5 cm long) flakes, which are called chips. These small 
flakes indicate stone striking activities on a site. 
Since the first appearance of stone artifacts 2,5 mio. years ago 
in Africa, the striking technique became step by step more 
complex. One example for a more complex striking technique is 
the hand axe, which appears for the first time 1,5 mio. years ago 
in Africa. The next big step in the development of the stone 
striking technique is the Levallois technique which is associated 
with the Middle Palaeolithic. The blade technique is linked with 
modern humans and is about 40.000 years old. Flakes and 
blades can be transformed by retouching into different tools like 
side- and end scrapers, knifes or points. Cores can also be trans 
formed by further treatment to core tools. Detailed analyses of 
flake and core surfaces give indications on the technical abilities 
of our prehistoric ancestors. 
3. 3D SCANNING 
3.1 General remarks 
As the objects themselves cannot be examined by all interested 
parties, very good digital representations have to be made 
available. In the case of the complicated stone artifacts con 
cerned, this demands a high resolution 3D documentation which 
comprises the complete surface which consists of sharp edges 
and larger areas which are relatively flat. Since enormous 
quantities of objects have to be documented, the method will 
only be acceptable if it works fast and efficiently. 
3.2 Selection of an appropriate scanner 
Both 3D scanners already owned by i3mainz, a Mensi S25 and a 
Leica Cyrax 2500, were not suited for the task since they are 
designed for much larger objects. Therefore, a new scanner 
featuring sufficient accuracy and resolution for small object 
spaces had to be acquired. The ATOS II system, produced by 
GOM mbH (GOM, 2003), was chosen from a large list of 
possible candidates with similar features for sub-meter object 
spaces (i3mainz, 2003). 
A main advantage of the system is 
its modular design which allows 
various fields of view between about 
1 dm 3 and 1 m 3 when certain com 
ponents (base, lenses) are exchang 
ed. For the stone age tools, the 
ATOS II is used in a version that 
allows to scan a measuring volume 
of 250 x 250 x 200 mm 3 . Fringes are 
projected onto the object’s surface 
(fig. 2b) and recorded by two CCD 
cameras which are located at both 
ends of a 600 mm base. The 
software computes 3D coordinates 
for up to 1.3 million object points 
from the camera images with an 
accuracy of about 0.01 mm. The 
scanner has a weight of about 4 kg 
and can be mounted on different 
types of tripods (fig. 1). 
Fig. 1: The ATOS II on a heavy tripod 
3.3 Scanning procedure 
First object position. The ATOS sensor head mounted on a 
tripod can easily be positioned relative to the stone age artifact 
which is fixed in special plasticine on a wooden block, where 
the adhesive ATOS reference points are attached (fig. 2b). 
During the measurement different fringe patterns are projected 
onto the artifact (fig. 2b) for some seconds each which are cap 
tured by two integrated cameras at either side of the sensor base 
(fig. 1). Within some more seconds the ATOS software calcu 
lates precise 3D coordinates of up to 1.3 million object points. 
Fig. 2a: A typical stone artifact. 
Fig. 2b: The artifact held in position by plasticine 
on the base block carrying the ATOS reference points. 
Fringe patterns are projected for measurement. 
During the scan the software checks the ambient illumination 
and possible relative movements between scanner and object 
because these can have an influence on the accuracy of the 
measurement. If one of these checks indicates problems, the 
software gives a warning and the scan can be repeated. Because 
of the critical influences of light and stability, the best place for 
measurements based on light pattern triangulation is a dark 
room with a solid floor. Surface reflectivity may cause 
problems, too. If the material is too shiny or transparent, the 
projected fringe patterns may not be identified correctly by the