TRY 
eth. An 
'riments 
üch we 
nologies 
s; using 
ression- 
ar single 
th both 
hat is, a 
pared. 
that the 
nced by 
the dies. 
on with 
>cted the 
ytogram- 
sted ito 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part BS. Istanbul 2004 
  
2. THE METHOD OF INVESTIGATION 
2.1 Producing the sample dies 
For the purposes of this investigation, we have devised and 
manufactured an instrument to produce laboratory conditions 
close to in vivo conditions and to standardize the impression 
taking procedure in vitro. The instrument consists of a platen, a 
part corresponding to the dental impression taking spoon, as 
well as a pair of rails with a sliding surface to keep the 
movement of the spoon onto the platen identical across 
instances (Fig. 3). 
  
Figure 3. The sampling instrument 
One of the most important features of an instrument of the kind 
shown above is its reproductive capacity, meaning in this case 
the precision with which it can be reset exactly to the same 
position while producing two different samples. Due to 
appropriately strict assembly and well-chosen production 
tolerance, as confirmed by an independent expert investigation 
(Budapest University of Technology and Economics, 
Department of Precision Mechanics and Optics), the zero 
position resetting accuracy of the instrument is 0,004 mm. A 
phase of the calibration of the instrument is shown in Fig. 4. 
  
Figure 4. The truing of the sampling instrument 
The platen of the instrument shown in Figs. 3 and 4 
accommodates one to three stainless steel dies that imitate the 
original abraded dies. These steel dies were made in a regular 
geometrical shape to facilitate comparison and, using a surface 
grinder, their surface roughness was made similar to that of 
abraded teeth. An important component of standardization was 
that the individual samples, produced one after the other, should 
be pressed against the platen with the same force. This was 
achieved by imitating the average human pressure of 20 N by a 
2 kg weight piece placed on top of the spoon. 
With this apparatus, impressions were made in 
four arrangements (single knife edge die, single shoulder die, 
three adjacent knife edge dies, three adjacent shoulder dies), 
with all four techniques. Thus, a total of 16 different models 
have been made; a 12-piece series was then made of each 
model. The impressions were then cast with Kromotypo 4 hard 
plaster; good quality setting was facilitated by using a vacuum 
mixer and a vibrator. 
The next task can be briefly summarized as 
follows. The various procedures will be assessed in terms of 
statistical correspondence between the set of hard plaster 
models and the original steel dies. The impression taking 
procedure for which the statistically demonstrable difference 
between the original shape and the model produced is the 
smallest will turn out to be the best. 
2.2 Determining the geometrical data of the model pieces 
For determining the parameters of the model pieces, we 
employed the method of photogrammetry; i.e., the data were 
not directly measured on the objects concerned but on 
photographs made of them. In particular, we obtained 
coordinates by a method now generally used in close-range 
photogrammetric tasks, an instance of which is the present task, 
i.e., by what is known as multi-viewpoint photography. For the 
calculations, we used a DLT (direct linear transformation) 
program developed at the Department of Photogrammetry and 
Geoinformatics of the Budapest University of Technology and 
Economics. The program and its testing procedure were 
reported on elsewhere (Detrekói 2002). 
In order to perform photogrammetric tasks, it is necessary that 
the pictures exhibit control points: it is with the help of these 
that the spatial position of our pictures is determined. It is a 
usual method in close-range photogrammetry that the spatial 
position of these points is not determined task by task; rather, a 
"test-field' applicable to a number of tasks is prepared 
beforehand. This was done in the present case, too. The dies 
were put into a test-field developed by the Department of 
Photogrammetry and  Geoinformatics of the Budapest 
University of Technology and Economics. The geometrical 
characteristics of the test-field were determined with a Zeiss 
Opton 3D coordinate-measuring instrument. That instrument 
shows the coordinates of points with a tenth of a micron 
accuracy and with a mean square error less than a micron. The 
calibration of the test field is shown in Fig. S. 
  
Figure 5. The calibration of the test field 
with a Zeiss 3D coordinate-measuring instrument