CALIBRATION FLIGHTS 1987
Due to bad weather conditions the test. range. was. only in
service in October 1987. Despite this three cameras could be
flown: for: calibration purposes on four -stripsc inca. crosswise
manner with each strip. flown'in bothndlirections.ciAstypitad
configuration of real camera positions with appr. 90 % longitu-
dinal lap is shown in Fig.1. Image scales. vary for-a' WA camera
from appr. ‘1 :3,5001 and) 166,000" and for:a NA! camera» from: appr.
124,000: and !1:5,300, according (toa. maximum’ elevation. differ-
ence in the range of some 480 m and flying heights above sea
level of 2,200 m and 2,900 m respectively.
Due to being very late in the year sun elevation was very low
which resulted in low illumination and long shadows.
MENSURATION OF PHOTOS
The last mentioned problems prevented almost errorless
measurements: off image-ccoordinates ats high precision
monocomparator. Instead of this they were performed at a ZEISS
Planicomp C 100 in four different positions for each photo for
the following reason. The photo coordinates of all points were
known a priori with good approximation, thus enabling automatic
prepositioning of the points and avoiding blunders. Despite
this precaution standard deviations of image';coordinatesuTe-
mained high.
FIRST CALIBRATION RESULTS
The requirements for a first overview were best met by a calib-
ration flight, dated 1.10.87, of Rheinische Braunkohlenwerke AG
of Cologne with a ZEISS RMK A 15/23 equipped with a Pleogon A2
of very low: distortion: Lowsflying height^ata'"cruise^speed: of
110 kn asked for an exposure time of 1/450 sec. Local time was
appr.:.10.30:hnrs. Thus: the shadowy: ‘parts of steep northern»slope-
es showed almost sno contrast on the original photos.) This led
to: pointing errors of appr. + 3 micrometers for the mean of
four pointings, including however also errors of the Planicomp.
Its comparator coordinates were reduced to photo coordinates by
affine transformation.
Tab.1 gives a selection of some interesting results from quite
a» lotsof/sdifferent: «calibration computations. Unvariaebly.the
following applies: Standard errors a priori for photo coordina-
tes are set at + 3 micrometers. Only points at the outer margin
(10 mm) of photos are given twice this amount, close to the
border however + 10 micrometers as well. Fig.1 shows among
others the taking position of the photos involved in different
adjustments.
The presented variations differ in the following characteris-
tics (cf. Fig.l):
1) 6round coordinates of 17 points are given in a Cartesian
system for x,y,z and for another 11 points for z only. from
32
