done with miniature cameras). Image quality is described in terms of
’’perception units”, such that each perception unit corresponds to a
just noticeable difference in quality between images being compared.
Heynacher’s proposal is that the logarithm of the average value of the
0. T. F. curve up to a limiting spatial frequency corresponds to the
number of perception units. For miniature camera work, the limiting
spatial frequency was taken as 40 lines per mm. and good correlation
was obtained with subjective quality rating. By weighting the average
0. T.F. towards the low frequency the correlation could be improved still
further. Dr Meier suggested that a modification could be used as a
quality criterion in photogrammetry. The limiting spatial frequency
could be set at 35 lines per mm., and the weighting function give more
weight to higher frequencies than in IIeynacher*s work. A practical
investigation of this idea would be most valuable.
Mr. Zorn. I. T. C. Delft, Netherlands described an instrument
developed at I. T. C. for testing the acuity of stereoscopic vision (see
Photogrammetria 20 (19^5)).
Mr. Hempenius. I. T. C. Delft, Netherlands described the use of
0. T. F. concepts in establishing the performance criteria for each
element of a photogrammetric system. A system of many components, e. g.
a photogrammetric system, has an overall transfer function that is
approximately Gaussian, and therefore has an approximately Gaussian spread
function. The combination of Gaussian spread functions is particularly
simple by adding variances. This idea was applied to the desirable size
of measuring marks, quality of stereoscopes, accuracy of edge location, eto.
Mr. E, Icelander. Sweden, described an experiment in which the
0.T.F. of a photographic system (camera, film, image movement) had been
calculated for an actual flight, and then compared with actual results.
Ground targets and well defined edges of natural objects had been used to
measure the actual 0. T.F. Agreement with calculation was found to be good.
4. Session 4 Sept. 22nd
Prof. B. Hallert. Sweden, talked on some aspects of camera
calibration. He queried whether the calibration procedures under laboratory
conditions recommended by Commission I were valid under flight conditions,
and he also thought the results were affected by the flatness of the plates
used in calibration, and of the film in the air. The reduction of observa
tions made during calibration was also important and the establishment of
tolerances for the errors likely to be met. He demonstrated these points
by comparing results of calibration by various methods.
Dr. II. K. Meier presented a paper by Dipl Ins. II. Ziemann,
Federal Germany, based upon a thesis to be presented to Stuttgart Technical
University. This covered a comparison of the distortions, measured with a
reseau, for cellulose acetate and polyester bases He had also evaluated the
improved accuracy obtainable using a reseau. Although the differential
distortion with polyester base was much less, the random errors were little
different for the two bases. The main use of the reseau wr-s to spot check
for distortion at a few points in each film strip, this procedure revealing
any large non-systematic errors.