vectors can be plotted in different projections (e.g. parallel 
projections from different sides of the sculpture). 
The results of both methods can later be combined for 
generating various visualizations of the sculpture including the 
digital reconstruction of vanished parts of the statues like the 
crown or the loincloth, which originally were made of wood 
covered with a layer of gypsum colored in gold. 
The small statue was recorded using the same techniques and 
procedures as with the bigger one. Problems occurred regarding 
the recording process: The statue is mounted onto a base of 
plexiglas and also it is fixed at the back with a plexiglas 
structure which could not be removed for the documentation 
process. As optical methods for recording are used, the 
refraction of the light passing through the plexiglas should be 
modeled for the data captured from the back of the statue, which 
seems to be anywhere between difficult and impossible. Thus, 
these data sets have not been processed until now. 
Figure 2. Registered point cloud from laser scanner. 
LASER SCANNING 
The statues were scanned using a MENSI S25 scanner. This 
scanner can be used in a range between 2 m and 20 m and can 
reach an accuracy of about 0.3 mm for the closest distance 
under optimal conditions. It is a triangulation scanner that sends 
out the laser beam at the one end of the scanner base and 
records the 3D position of the reflected point using a digital 
camera at the other end. The base for this scanner is about 80 
cm. The opening angle in this plane is about 45°. Additionally, 
the scanner can rotate around its horizontal axis and in this way 
has a vertical opening angle of 320°. The accuracy of a point 
measurement is dependent on the distance to the object due to 
the triangulation concept of the scanner. The scanner can 
measure with a rate of 100 points per second at most. 
As the calibration of the scanner was not optimal for closest 
distances at the time of recording the sculpture, the accuracy of 
the derived surface model can be expected to be about 0.5 mm. 
This should be sufficient for measurements to be taken at the 
model, e.g. the distance between certain points on the statue, but 
this model does not allow an accurate reconstruction of the 
sculpture, which was not intended in this case, anyhow. 
One challenge in scanning complex 3D objects like this statue is 
to cover the complete surface with the scanning process. This is 
supported by software tools allowing the visualization of the 
scanned point clouds, usually supplied by the scanner’s 
manufacturer with the software controlling the scanning process 
itself. It is highly recommended to do further checking by 
triangulating the surface to visualize possible holes that are 
often not easily to recognize by just inspecting the point cloud. 
The process of scanning the bigger sculpture took about 6 days. 
The working hours of the single days were short due to the 
opening hours of the museum and the fact, that the scanner was 
not allowed to be operated unattendedly during night time. In 
this time, altogether about 1.8 million points on the surface of 
the statue were scanned from 29 different observing points. The 
scans were performed with a mean point grid on the surface of 
the statue of about 1.0 mm for every single scan. This point grid 
is densified considerably as the surface is usually covered in 
multiple scans from different directions. With regard to the 
accuracy of the scanner and the time for scanning, an even more