nines the 
age co- 
itor was 
s of our 
ition was 
eoplotter. 
les is a 
in case 
nalytical. 
à spatial 
a general 
sofwear. 
nage co- 
n up the 
ow of a 
made a 
rates and 
der lines 
ering in 
zation of 
ing and 
a Kodak 
G 
F bundle 
nage co- 
-ordinate 
ordinates 
f control 
splay in 
nt result, 
meras so 
previous 
ordinates 
v frame. 
tes of the 
a were 
culations 
s we had 
tation of 
nological 
> applied 
first case 
> control 
e to the 
uncertainty of the position of the three 
planes comparing them to each other. 
For this reason in the case of several 
points in the same plane we did not 
bring co-ordinates but only fictive 
distances into adjustment. In the second 
case central points marked by a cutting- 
off lathe on an iron cylinder plane 
surface produced by a surface grinder 
were considered control points. Their 
plane co-ordinates can be determined by 
a measuring microscope and their height 
values by a micrometer of a mean square 
error below 0,01 mm. In this case the 
co-ordinates of the points can be brought 
into adjustment like vxact values. 
  
Figure 3.: The contour lines of the 
upper gums 
5. ANALYZATION OF THE 
RESULTS 
As it has Leen clearly demonstrated, we 
produced image co-ordinates from each 
image and the adjustment was made 
with the same programme. The only 
exception was the contour line of a 3 
mm contour interval of the human gums 
produced by an analytical plotter (figure 
3.). During the computation our best 
results were gained by using additional 
parameters; the explanation of this fact is 
that the inner orientation were unknown 
and this is a demand for displaying of 
other parameters in this area like 
distortion and film motion in the 
165 
formation of the funtional modell. This 
is true even in the case when we do not 
want to co-ordinate physical content to 
additional parameters. We put the 
summary of our results in a table. Every 
value is given in respect to the image 
space in order to make easier 
comparisons to other results; everyone 
was made with the same objective and in 
the same photographical arrangement. 
m uM Mas oc UMgsof imi 
1 30 23 16 20 
2 2,5 19 17 19 
3/552,553,0:5:022 16 23 
S Sinis 18 24 
4 
where 
1 -Nikon film, 
2 -Kodak DCS, 
3 -Dicomed, 
4 -Nikon dig., 
m,  -isthe mean square error of 
image co-ordinates brought into 
adjustment, 
Mor -is the average mean square 
error of the object space co- 
ordinates previous to adjustment 
in the image space, 
Hoo -is the average resultant square 
error of the object space co- 
ordinates after adjustment in the 
image space, 
ll,  -is mean square error resulting 
from the differences between the 
co-ordinates of known points but 
which was not drawn into the 
determination of the parameters 
of adjustment and the co- 
ordinates gained from the 
photogrammetric point 
determination of the same points. 
It can be laid down as a fact that using 
the methods of close-range 
photogrammetry it is possible to solve 
the task with the accuracy needed. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996