(175;
plotted as a function of the focal setting which provided the minimum con
trast threshold value for that particular size of symbol.
One of the major projects over the past years at our laboratory has been
assessing the criteria for recognition and detection of photographic detail. The
techniques of this type of experimentation have been previously discussed 7 ’ 8 .
If the significance of the experiment may be compressed in one sentence, it
might be stated that we hold that photographic quality in fields of technical
photography is determined by the amount of information which the photo
graphic interpreter can extract from the picture—that, given two photographs
of the same object, the one which enables the photographic interpreter to
detect more symbols is the better quality photograph. The second experiment
of interest in today’s discussion was conducted with this same lens to assess the
probability of detecting symbols on the photographic material. For this ex
periment, simulated one-foot objects were photographed at various scales.
Scale was then plotted as a function of the focal setting which provided the
maximum probability for detection of the symbols. The results of the two
experiments are compared in Figure 6
wherein good agreement is evident
between the focal setting which pro
vided minimum contrast threshold for
a given image size, and that which
provided peak probability of detection
of the corresponding size of symbol.
We therefore conclude that, be
cause the threshold characteristics in
tersect as focal setting is changed, as
shown in Figure 4, and because of the
establishment of the fact that the focal
settings for minimum contrast thres
holds and peak detectability are in
agreement (Figure 5), (a) any judg
ment of relative performance of pho
tographic systems based purely on a
resolution number derived from a
high-contrast test object may well be
misleading, (b) any judgment of oper
ational focal setting of the photo
graphic system based purely on a reso
lution number derived from a high-
contrast test object may well be incor
rect. More meaningful approximations
have been suggested with an eye towards simple laboratoiy pioceduies, howe
ver, judgments based on any resolution or sharpness considei ation deal only
with a single size of symbol.
Fig. 6. Focal Settings as a Function
of Scale.
Our interests in airborne operation deal with a range covering man) siz^s
of symbol. For this reason, an absolute rating which must be established should
be based on a function which weighs the quality over the range of symbol
contrasts and sizes subject to consideration. The area over the contrast thresh
old characteristic of the system is suggested as being this type of function.