International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004 Intei 
Four samples were put into the test-field at a time, parameter at hand. In determining that technological value, we 
corresponding ones always to the same place, before the used error filtering based on statistical tests where required; 
photographs were taken. With this solution, we made sure that then the arithmetical mean of the filtered values was accepted 
results were not distorted by differences in geometrical as the final value. The parameters of the metal dies, i.e., the 
arrangement. The photographs were taken with a Nikon DI technological values of the various methods of impression 
2000X1320 pixel digital camera, from the same four viewpoints taking, were compared with respect to bottom diameter, middle 
in each case. One of the pictures used. for subsequent diameter, height, overall width, and, for shoulder arrangements, 
calculations is shown in Fig. 6. shoulder diameter. Numerical values of the results are shown in 
Figs. 8, 9, 10, and 11. Adding all differences measured in the 
rarious test dimensions, we get a single value, call it total error, 
à that characterises each procedure investigated. This is shown, 
d along with typical distortions found with plain dies (meaning 
| ones without either place-holding or grooves), in Figs. 12 and 
| 13. Average values of bottom and middle diameters of knife- 
edge model dies compared to those of metal dies for the four 
impression taking methods (in mm) 
Ave 
thos 
mm 
Figure 6. A picture used for subsequent calculations 
The pictorial coordinates were measured using a PC, with the 
help of software packages of a sub-pixel accuracy level. 
3. RESULTS AND DISCUSSION 
3.1 Determining the geometrical parameters 
The objects symbolizing the original dies were regular 
geometrical solids; in particular, they were composed of a 
truncated cone sitting on a cylinder. In the case of the 
‘shoulder’ arrangements, the latter was mounted on another Figure 8.a Bottom and middle diameters 
cylinder of a longer radius. The reason for this was that fewer 
parameters were required for a geometrical comparison of the Average values of bottom and middle diameters of shoulder 
original object and the copy than would be the case with model dies compared to those of metal dies for the four 
irregular shapes. The parameters used numbered ten on average, impression taking methods (in mm) 
across the four different methods of impression taking; for Avi 
stand-alone dies that number was lower, whereas for sets of con 
dies, it was higher. These parameters represented the main mei 
dimensions of the given solid like bottom diameter, middle 
diameter (i.e., diameter of the circle where the cylinder and the 
truncated cone meet), height of die, distance between dies, 
overall width of dies, diameter of circular shoulder, etc. Since 
photogrammetry is a method of determining spatial coordinates 
of points, the parameters were distances between pairs of points 
in all cases. For the identification of those points, we milled 
marks on the metal dies, faithfully mapped on each plaster die. 
Fig. 7 shows the parameters selected. 
insimul 
Figure 8.b Bottom and middle diameters 
Average values of height of knife-edge model dies compared to 
those of metal dies for the four impression taking methods (in 
mm 
Figure 7. Parameters measured in the experiment Av 
Cor 
me 
  
   
With the application of that method, we got a twelve-element 
sample for each selected quantity; the value composed of those 
values was then accepted as the technological value of the