18
In this test a grid model was formed to simulate a frame photo
graph model. The coordinates of each grid intersection were analyti
cally deformed using a set of non-conformal third degree polynominal
distortion equations in X, Y and Z. The coefficients of these poly
nominal equations were fed as constant data into the Black Box
Program and the instrument deformed model, was compared to the analyti
cal deformed model. Figure 8 shows the magnitude of deformed data
used in this test. As you can see, large distortions were introduced.
Figure 9 shows the residual model distortions remaining after the
Black Box adjustment of the model to the theoretical distorted values.
The unit standard error of residuals after adjustment was:
a x = 11.9b
= 11.5m-
a z = 23.0m-
12.2 Black Box Adjustment of Panoramic Materials
The computational precision test proved the capability of the
AP-2 to compensate for extreme non-linear distortions. An interesting
application of this capability was to remove the distortions from a
stereo model constructed using stereo pan materials taken at a
comparatively high altitude over Shaw AFB, Alabama, The camera employed
for these tests was the KA 52 panoramic system having the following
characteristics :
f
Scan angle
Format
Cycling ratio
Resolution
= 75 mm
= l8o°
= 4 l/2" x 12"
= 6 cycles/sec.
=70 lines/mm
Two exposures having common overlap and a convergent angle of
30 0 were selected over a well mapped area in the vicinity of Shaw AFB.
After setting up a 2x enlarged portion of the stereo pan model in
the AP-2 and orienting it absolutely to map control, the residuals
shown in Figure 10 at model scale were obtained. Several adjustments
were made to determine the best set of coefficients for the Black Box
Program. The final set of non-conformal equations selected were of
the form: