43
Regular errors of the image coordinates known from camera tests under operational conditions have
been corrected and corrections have also been determined from measurements of the fiducial marks of
the actual photographs. These corrections have been included in the program. All image coordinate
measurements were made in the sterocomparator Wild StK 824, see 2.1.
For the relative orientation, image coordinates have usually been measured in 15 points well distri
buted over the surface of the model. In addition, the coordinates and parallaxes were measured in all
signalled points of the model. Five points, one in the center and four in the corners of the model were
used for the absolute orientation.
From the computations the standard error of unit weight of the y-parallax adjustment, the residual
y-parallaxes in all points, all elements of the relative and absolute orientation, and the final transformed
coordinates of all signalled points were determined. The discrepancies between these coordinates and
the given geodetic coordinates were determined and expressed as root mean square values in X, Y, and Z.
According to previous investigations (see reference 1.342:1) there is a theoretical relation to be
expected between the accuracy of the relative and absolute orientation procedures on one hand and the
final coordinates on the other. The standard errors of the final coordinates can be expressed in formulas
as functions of the standard errors of unit weight of they-parallax measurements and of the model coordi
nate measurements. The number and the location of the control points will have some influence upon the
formulas. For five control points in planimetry and elevation, located in the center and in the corners
of the model, the following expressions have been derived:
h , 1/1 *\ 2 2y 2 \b x(x—b)\ 2 1 (x(x—b) 2
s * = 7 s ° p \ (t - t) w + \20 + ~2<P + j b
M 2 166 2 d 2 -6 4 | 2 1 I bd 2 x(x—b)\ 2 3y 2 ,
' b 2 + 4d 2 + 40(6 2 + 4d 2 )j b 2 d 2 + (6 2 +4d 2 + b j 4d 4 +
— 6) | \x(x—6) 2 xbd 2 16b 2 d 2 — 6 4 | 1
b | j b b 2 + U 2 + 40(6 2 + 4d 2 ) | bd 2 +
'20^^20b^20f r ^f2W~Y96d 2
4 20(6 2 + 4d 2 ) * s' 2 op
h , |/ 2(4d 2 — 5y 2 ) 2 (2x-b) 2 y 2 \x 2 3x b(b 2 + 2d 2 )] 2 y 2
7 S op \ iw" 1 Wd 2 4 (b “ 2 1 2(b 2 + 4d 2 )J b 2 d 2 +
( xy 2 y 2 (2x—b)bd 2 j 2 3 (2x—b)y 2 (x 2 3x (
. _ + — + 2 (6 2 +4d 2 ) ( 4d 4 4 2b 2 d 2 |7 ~ ~2 ~ r
b(b 2 + 2d 2 ) | , 20x 2 - 20bx + 20y 2 + 29b 2 + 96d 2 ¡77
+ 2(6 2 + 4c? 2 )J + 20(6 2 + 4d 2 ) ‘ s /2 op
(A +
(20 +
x(x
h? , 1/1 {/ ** by 3(2*-6)V| 6 (2x-b)‘ y*
be S op M 2 b X 20) ! 16 d 2 j 1 5 + 44* 4 d 2
The symbols in these formulas are:
s x , s y and s z the standard errors of the final x, y and z-coordinates on the ground
s' op the standard error of unit weight ofy-parallax observations on the scale of the photographs
s'oc the standard error of unit weight of image coordinate observations on the scale of the photographs
h the flying altitude
c the camera constant (image constant)
b the base
d the distance from the base to the lateral orientation points
x and y the model coordinates on the same scale as b and d. The origin is located in the left nadir
point and the positive #-axis along the base.