of reference that gives the placing of the transformation in table 2. The num-
bers of table 3 are microns in the scale of the photographs. The last column
at the right gives the root mean square value of the residual errors of each
photograph. The last row of the table is composed of the corresponding mean
value of the transformation in question, which expresses the degree of accuracy
that has been reached with the transformation. This value shall be refered to
as the "accuracy factor" of the transformation in the following.
The detailed information of table 3 enables the reader to gain a thorough know-
ledge of the deformations that have taken place, at least to the extent that
it is possible to portray the deformations on the basis of the positions of
the four fiducial marks. C e
An abridgement of table 3 (table 4) contains only the symbol of the transfor-
mation followed by a column that gives the number of the photograph and fidu-
cial mark that has the largest coordinates residual error and a column
that contains the accuracy factor of the transformation.
Analysis of the results of the transformation calculations
The original film negative, and the diapositives that have been prepared of it
have been made in Czechoslovakia in 1966. The measurements pertaining to this
material were carried out in Finland in June of 1967 and repeatec about half
a year later in December of 1967. The above mentioned measurements of the re- I,
lative position of the fiducialmarks of the camera have also been at our dis- ‚> €
posal.
It can be said of the reasults obtained by affine transformation that the
accuracy factors of the transformations vary between 44 - +8 microns while
the maximum values vary between 7 and 20 microns. These values can be consider-
ed fully adequate and thus it can be said that the material that has been pro-
vided to the participants for the experiment is fully up to the standards re-
quired for_analytical methods.
The planned experiment was however intended to be carried out in an analogy
instrument that generally does not offer the possibility for affine transfor-
mations, in which case the possible deformations of the film base correspond
to the results given by Helmert transformation.
