The standard errors of the coordinates x and y are then obtained as
y
—
m, = Sy J Qu (38)
m, ES $9 y, (39)
sy is the standard error of the image coordinate measurements (with
minor approximation). The expressions (38) and (39) are graphically
demonstrated in the diagrams 3 and 4 for s, — 1.
The magnitude of s, depends upon a great number of factors, as for
instance, the negative material (film or plates), the quality of the lens,
the accuracy of the actual measuring device ete. For wide angle film
cameras c = 150 mm and high precision measuring devices the magnitude
of s, is about 0,01 mm.
In the same way the error propagation and distribution can be in-
vestigated for other positions of the control points. In diagrams 5 and
6 the distribution of the standard errors is demonstrated for another
position of the control points. If more than two control points are given
there will be discrepancies in the control points in connection with the
coordinate transformation. In such a case some condition must be
specified for the adjustment of discrepancies. The simplest and most
convenient condition is that after the transformation the square sum of
the discrepancies in the control points shall become a minimum. After
introducing this condition and treating the discrepancies as functions
of the basic measurements the error propagation and distribution can
be determined as above. This method was used by HALLERT 1956.
1.16. Systematic errors of the intersected coordinates due to inclinations of the
photographs
We shall assume the image coordinates to have been measured in
coordinate systems, the origins of which are located in the principal
points. The intersections were then computed according to formulae
(1) and (2) above. If, however, the inclinations of the photographs were
94 €, Q9 05, respectively, and if the elevation differences on the ground
were small in comparison with the flying altitude, the origins of the
image coordinate systems should have been chosen in the isocenters of
the photographs, see fig. 3. The coordinate systems, the origins of
which are located in the principal points, are denoted (x')(y') and
(x") (y") respectively. The corresponding coordinate systems, the
origins of which are located in the isocenters, are denoted z'y' and
x" y" respectively. The image coordinates of the isocenters are
9