method for calibrating aerial cameras was not considered seriously
for some time. By experimenting with a number of filter and light
combinations, however, we found that we could approximate the cali
brations obtained with the field range and multicollimator if the
target plates were adequately illuminated. A combination of No. 73
Wratten filter and a flourescent light source provided optimum
results.
TEST PROCEDURES AND RESULTS
CORRELATION
Since lens calibration data obtained from the goniometer when
using the fluorescent light source and a No. 73 Wratten filter appeared
to closely approximate the calibration data obtained from photographic
methods, we conducted a series of tests to see how closely the two
methods correlated. Camera lens Type T-ll, No. XF-2256, was selected
for the tests. Closely correlated calibration data already were available
on this lens from the Fairchild Camera and Instrument Corporation's
multicollimator and from the field calibration range. Results of
these calibrations are presented in Table 1.
When comparing the results of calibrations in Table 1, a maximum
deviation of 10 microns appears between the radial distortion curves
for the goniometer and the multicollimator. The maximum deviation
between the curves for the field range and the multicollimator is
3 microns.
REPEATABILITY
Further tests were conducted to determine whether the same precision
could be obtained with repeated tests. Data was first recorded when
using the No. 73 Wratten filter and a tungsten light source. Results
of each of the calibrations were tabulated, and curves were drawn from
an average of the four diagonal halves for each. These curves show that
75 percent of the targets fall within +1 micron and 94 percent within
+3 microns of the average deviation. The remaining 6 percent of the
targets, measured at angles of 45°27' and 46°40', fall within 5 to 6
microns of the average.
Similar calibrations were made on Type KC-1 camera lens cone
No. XF-6765, using the goniometer with a No. 73 Wratten filter and a
fluorescent light source. Four individuals recorded the results of
seven observations. Plots of these readings indicated that 76 percent
of the points fall within +1 micron of the mean and 97 percent within
+3 microns. In addition, 84 percent of the readings of each diagonal
fall within +3 microns of the mean. The adjusted average distortions
for each observer and the mean for the group are presented in Tahle 2;
the mean is plotted as a single distortion curve in Figure 4, as com
pared with the multicollimator calibrator.