GEMS STUDY PROGRAM SUMMARY FINAL REPORT

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B04747A000700010025-7
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 25
• Document Creation Date: December 28, 2016
• Document Release Date: March 29, 2002
• Sequence Number: 25
• Publication Date: October 20, 1967
• Content Type: REPORT

Approved Fo 'F # ? c 22/06/17 : CIA-RDP78B04747A000700010025-7
Declass Review by NIMA/DOD
25X1A
Approved For Release 2002/06/17 : CIA-RDP78B04747A000700010025-7
S--I Ei vsH - M - OOW-3
 ENGINEERING REPORT NO.
GEMS STUDY PROGRAM
SUMMARY FINAL REPORT
DATE: October 20 1967
PREPARED FOR: SPO 27203
2X1 A
0
Approved For Release
p-oved For. Release 2002/06/17 CIA-RDP78B04747A000700010025-7
'JSE-,tc\X-COOS ..3
 Section
Approved For Release 2002/Q6/17 :LLCIA-RDP78B04747A000700010025-7
TABLE OF CONTENTS
GEMS STUDY PROGRAM
1.1 Introduction
1.2 Program Objectives
SUMMARY OF TASK RESULTS
2.1 Refinement of. Techniques
2.2 Alternate GEMS Technique.
2.3 System Parameters Study
2.4 Equal MagnificationGEMS Study
2.5 Psychophysical GEMS Study
2.6 GEMS Viewer Study
2.7 P,,rudo GEMS Viewer Study
PROGRAM CONCLUSIONS
3.1 Technological Status
3.2? Usefulness of GEMS
3.3 Program Recommendations
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
Page
 Approved For Release 2002/06/17 CIA-RDP78BO4747A000700010025-7
ACKNOWLEDGEMENTS
The GEMS Study program could not have been successfully completed
without the assistance received from various customer facility personnel. The
author wishes to thank all the individuals who contributed their time and
efforts to the Psychophysical CEMS experiment.
A special expression of gratitude is extended to
who 25X1 A
did a fine job of coordinating customer-
25X1A
aided us in, acquiring the system parameters information and conducting the
psychophysical experiment.
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
GEMS STUDS PROGRAM
1.1 INTRODUCTION
The GEMS Study program has been successfully completed. The program
efforts succeeded in accomplishing the major objectives. Techniques were
established for generating realistic simulations of the appropriate mission
material; the accuracies of the simulation processes were defined for most of
the simulation technique; and the psychophysical study was effective in defi..n-
ing the increment spacings for a modulation transfer function (MTF) - ground
exposure GEMS matrix array and the accuracy of the subjective judgments of
each parameter.
Since a comprehensive final report was submitted at the completion
of each program task, the program final report wil" serve to summarize. the
findings of the tasks and to discuss the potential. implementation of GEMS in
the evaluation of mission material.
1.2 PROGRAM OBJECTIVES
The GEMS Study program was concerned with seven basic tasks. The
major objectives of each task are defined in the following subparagraphs:
1.2.1 Refinement of Techni_gues
In refining the simulation techniques of the GEMS modified
contact. printer, the primary objectives were to improve the precision with
which. each of the imago quality parameters could be controlled and to determine
the accuracy of the simulation processes. The simulation processes and instru-
mentation were up-dated in accordance with the study findings. This task was
Approved For Release 2002/06117 CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06117: CIA-RDP78B04747A000700010025-7
later modified to include the study of MTF simulation control in the Pseudo
GEMS Viewer concept.
1.2.2 Alternate GEMS Technique
In simulating MTF with the modified printer GEMS instrument,
it was conceived that the control of this simulation parameter would be Fresnel
diffraction limited at approximately 100 cycles per millimeter. In the antici-
pation of future simulation requirements above this upper spatial frequency
limit, an alternate means of controlling MTF was investigated. The alternate
technique investigated involved the use of a modified copy camera system where
the spread function of the imagery was modified by controlling the size and
shape of the copy lens pupil function.
1.2.3 System Parameters Study
Under the redirection of efforts phase of the program, the
System Parameters study was initiated in order to obtain objective measures
of such appropriate mission material parameters as MTF, ground exposure, and
scene contrast; and such film-processing characteristics as the film sensito--
metric curve, film granularity, and film density neutrality. These objective
measures were used in establishing the parameters of the psychophysical GEMS
matrix,
1.2.4 Eqp alMagnification_GEMS Study
Under the redirection of efforts phase of the program, the
Equal Magnification GEMS study was established for the purpose of defining a
technique that would permit realistic simulations to he generated at the same
scale factor as mission material. This task evolved because ordinary negative
material could not be employed in generating a GEMS master transparency with a
Approved For Release 2002/06/17: CIA-RDP78B04747A000700010025-7
 Approved For Release 2002/6,6i177-: !CIA- 2DP78BO4747A000700010025-7
faithful reproduction of. the ground exposure tonal scale. In addition, the
control of the MTF simulation was influenced by the 'non-linearities in the
toe and shoulder areas of the original film characteristic curve.
1.2.5 Psychophysical GEMS Study
The major objr.ctive of the Psychophysical GEMS study was to
determine the usefulness of GEMS in the subjective evaluation of mission
material. The objective was to be accomplished by psychophysical experimen-
tation. The information to be acquir(,d'from this experimentation was to
a) the visual perceptible increment spacing of MTF.and
exposure,
b) the accuracy -of the subjects' judgments of each parameter.,
c) the uniqueness of each M`L1?-exposire .element in a matrix
array, and
d) a measure of the subject's ability to rank mission material
with GEMS.
1.2.6 GEMS Viewer - llesj_gn_Concept
If GEMS are to be employed in the evaluation of mission
material, it is essential that a viewer be designed for handling both the GEMS
and the mission material. The design features of such a viewer were established
under this task. Consideration was given to such features as an automatic
matrix element search mechanism, split field viewing, GEMS rotation, and vari-
able magnification viewing.
1.2.7 Pseudo GEMS Viewer
It was envisioned that' a GI-MS viewer could be constructed
where the parameters of MTF, ground exposure, and cease con.tias would be
Approved For Release 2002/06/173. CIA-RDP78BO4747A000700010025-7
 y '1
Approved For Release 2002/0647 : CIA-RDP78BO4747A000700010025-7
simulated by optical and'electrical.mechanisms within a viewer. The fabri-
cation of a viewer of this nature would eliminate the need of a large film
matrix array. This task was directed toward demonstration of the feasibility
of this concept with a breadboard experiment.
Approved For Release 2002/06/17't CIA-RDP78BO4747A000700010025-7
 Approved For Release 20Q2/06/17 : CIA-RDP78BO4747A000700010025-7
SECTION II
SUMMARY OF TASK RESULTS
2.1 REFINEMENT OF TECIUNIQUES
The Refinement of Techniques was concerned with investigating the
following simulation processes and improving the simulation instrumentation.
a) M1'F simulation on the GEMS modified contact printer
b) MTF simulation in the Pseudo GEMS Viewer
c) Exposure simulation
d) Haze simulation
2.1.1 Simulation Technique Conclusions
An analytical investigation was performed to determine the
effects of near-field diffraction on the control of MTF when the MTF was
simulated with the GEMS modified contact: printer. The results of the inves-
tigation proved that the simulation of MTF could be suitably controlled below
100 cycles per millimeter. Above 100 cycles per millimet_r the Fresnel diffrac-
tion effect introduces a sizable alteration in the shape of an edge gradient.
The edge shape alterations are too extensive to allow the simulation of specific
transfer function shapes in the high spatial frequency regions.
In the simulation of MTF with the Pseudo GEMS Viewer, non-
linear film imagery must be convolved with the degrading linear spread function
elements of the viewer optics. An analytical investigation was performed to
rictermi_ne the error that would result in the simulation of M F when employing
a negative scene as a master transparency. The investigation demonstrated that
a 160 cycle per millimeter MTF simulation could be achieved with less than a
4 percent error for the-full tonal range of the imagery. Errors of less than
5
Approved For Release 2002/06/17: CIA-RDP78BO4747A000700010025-7
 Approved For Release 200 06117 : CIA-RDP78BO4747A000700010025-7
2 percent would be introduced in an 80 cycle per millimeter MTF simulation.
Percentage errors of this magnitude are quite tolerable.
Modification of the GEMS instrument with a well regulated
light source has yielded exposure simulations within the tolerance limits of
the evaluation equipment. A 0.02 density exposure shift can be simulated.
An analytical and experimental investigation was initiated to
determine the accuracy of the. haze simulation process. Due to the film effects,
created by the double exposure process, haze simulations cannot be achieved
with less than a 5 percent error by the present technique. By modifying the
GEMS instrumentation. with a non-image forming, d.c., light source, both the
fogging exposure and the image exposure can. be accomplished simultaneously; and
the film effects would be eliminated.. Simulating haze in this fashion is
identical to a real haze situation except for shadow area back-scatter. The
error of the simulation process then would be reduced to the error of the
evaluation equipment. It is important to note that to obtain an estimate of
haze with GEMS is not dependent upon the phenomena of shadow area back-scatter.
2.1.2 Instrumentation Conclusions
The GEMS instrument was modified with a film vacuum platen
and a strobe flash unit.. The vacuum platen and its associate jigging pc,_-mi.t
control of the master transparency-GEMS film separation-to : 0.0002 inches.
At a MTF simulation level of 50 cycle:; per millimeter, the instrumentation
variables for the modified contact printer can be controlled to + 2 percent of
tin: desired transfer function. At 100 cycles per millimeter the error increases
to sliglitly less than + 4 percent.
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 : CIA-RDP78B04747A000700010025-7
A strobe flash unit was installed to-improve exposure repeat-
ability and to eliminate the film Reciprocity Law Failure problems introduced
by long exposure times. With the new source, an exposure repeatability of
+ 0.01 density units is achievable. A + 0.01 density variability is equivalent
to the error of the densitometer used in the exposure evaluation process. It
is quite possible that the source has less variability than can be detected.
with the evaluation equipment.
An aerial image read-out device was breadboarded in order to
obtain more accurate measures of sensitometric data in a copy camera reduction
system. In a copy system an error is introduced in the edge-gradient analysis
sensitometr.i.c data if the lens fall-off properties are not taken into account.
To obtain valid sensitometric data, sensitometric step readings must be obtained
in the lens image plane. The breadboard device provided sensitomet:ric data
within + 0.01 density units for the low density region and within + 0.02 density
.units for the high density region.
2.2 ALTERNATE GEMS TECHNIQUE
The present use of a modified contact printer for the simulation of
MTF provides predictable and controllable results over the spatial frequency
range of 0 to 100 cycles per millimeter. It is desired to extend the MTF
simulation capabilities to a higher frequency range; therefore, an analytical
investigation was conducted to determine the validity of several alternate
simulation approaches.
The alternate procedure established to be most valid involved the use
of a modified copy camera system with a variable--transmission, spread function
mask employed in the pupil plane of the lens for control of the image spread
Approved For Release 2002/06/17- : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 CIA-RDP78B04747A000700010025-7
function size and shape. A breadboard unit was constructed and some experiments
were performed to demonstrate the feasibility of the. simulation concept.
The investigation was considered to be highly successful. in view of
the fact that the only limitation was in the optics and not in the theories or
techniques utilized. The optics problem can be easily remedied. Despite the
limitation of the optics. the task produced several highly significant results.
First, the validity of the technique of MTF modulation by aperture
apodization of a copy camera was conclusively demonstrated. That the technique
was not demonstrated for spatial frequencies out to 200 cycles per millimeter
as intended was, as stated, the fault of the optics used, not. of the technique
itself. The design of better optics will allow extension of this technique
without serious difficulty to frequencies of 200 cycles per millimeter and
beyond.
Second, the good IMF agreement obtained between the "theory" and
experiment is an excellent demonstration of the repeatability and accuracy of
the edge-gradient analysis technique itself; although this was supposedly not
at stake here, it is a worthwhile corroboration of the method.
Third, the experience gained by encountering and partially surmount-
ing the breadboard difficulties is valuable for future efforts.
2. 3 SYSTEM PAI:AMETEPS STUDY
The GEMS Study program is primarily concerned with the simulation of_
a particular class of aerial photography. The realism achieved in the simula-
tion of this photography is dependent upon both the simulation techniques and
a knowledge of the system parameters. Various talks with the customer`s people
Approved For Release 2002/06/17: CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17" CIA-RDP78BO4747A000700010025-7
indicated that no objective measurements existed which would describe the
mission material parameters in a manner directly related to the objective
Under the redirection of program efforts, a task was established for
the purpose of objectively defining both the film characteristics and the
system performance factors. The goal. of the study was to define the following
parameters in suitable terms, for purposes of scaling and evaluating the GEMS
matrices:
a) film sensitometric curve
b) film granularity
c) film density neutrality
d) system modulation transfer function" (ylTJ?)
e) aerial. image contrast
f) scene exposure
The parameters of the study were defined in terms of objective
25X1A
measures wlr ch were apropos o the evaluation of GLMS and the use of
25X1A
facility photographic evaluation instrumentation.
The study data cannot be. condensed to single numbers without losing
significant information. For this reason; the data compiled in the text of
the. task report should be consulted for specific. details.
To abstract summary measures pertaining to system performance factors,
it can he stated that the study indicates the average aerial i.n-mage contrast to
be 0.274 rnodulati_on and the average scene exposure range to be a 0.44 density
difference for the si.x scenes evaluated. Figure 1 presents a co .posite gr..aph
of The measured system transfer functions.
9
Approved For Release 2002/06/17 : CIA-RDP78B04747A000700010025-7
 tl i~ $ t I t 1'
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
10 20
L1 p oII II n
o
^
~ ~-~ * p+ A+o Q+o .+OQ +0 + II
}
Q
! 1 1 I + 1 1' E i !~[ ~ 1~ I t 1
30 40 50 60 70 SO 90 100 110 120 130 140 150 160 170 180,
SPATIAL FREQUENCY - CYCi MM
Figur I - Scaled MTF Determination Composite
Approved For Release 2002/06/17 CIA-RDP78BO4747A000700010025-7
0.8r
SCALED MTF DETERMINATION
2 A
CORRESPONDING TO EDGE NO:
5 +
(Results Obtained from the
6 0
Evaluation of Mission Material)
7 0
8 x
10
0 .3 a-o ' o A
A+0 0
0.4 -
 Approved For Release 2002/06/17: CIA-RDP78BO4747A000700010025-7
The film properties provide a knowledge of.the film response and
imagery appearance. This data, as compiled, is to bo used to ensure proper
simulation of processing as well as-to furnish a means of monitoring all vari-
ables which affect film granularity and density neutrality.
The System Parameters Study can be termed quite successful in that
the data obtained served to define the operational level of a system whose
performance is rated as "good".to,"best." The data made available by this
study was inclusive enough to give guidance in specifying and evaluating the
various system parameters as related to the production of a psychophysical GEMS
2.4 EQUAL MAGNIFICATION GEMS STUDY
Earlier. GEMS efforts established the fact that the appropriate pro-
gram mission material could not be simulated from an original negative that
was processed to a high gamma-and/or photographed under conditions that intro-
dosed either a substantial degree of atmospheric haze or a compressed scene
exposure range. The Equal Magnification GEMS study-was initiated to investigate
not only the properties that a transparency must possess to yield a valid
simulation, but also the proper technique of performing the simulation. The
study treated all phases of the simulation topic. It clearly defined the
system parameters to be simulated and used these parameters as guides in
developing the overall aspects of a total procedure which would yield realistic
simulated photographs.
The study results concluded that to properly simulate MTF,exposure,
and contrast, requires both a unity gamma, positive transparency with a large,
linear film density?-exposure range; and?a modified copy camera for simulating
M'fr's above 100 cycles per millimeter. A positive transparency of the nature
Approved For Release 2002/06/17: CIA-RDP78B04747A000700010025-7
 Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
described above is necessary in the simulation process, because the accuracies
of the simulations are dependent upon the modification of imagery which is
linearly related to the exposure domain. A modified copy camera is essential
in order to predictably control MTF and to avoid the effects of'Fr.esnel
An experimental film study established that imagery obtained with
DuPont SR-102 negative reversal film would satisfy'allthe positive transparency
requirements.
2.5 PSYCHOPHYSICAL GEMS STUDY
A psychophysical study was conducted to determine the usefulness of
GEMS as a subjective evaluation tool. The study results definitely established
the fact that GEMS can be employed-to obtain 'realistic estimates of system
performance. The study experiments provide.d.valuable'_i.nformation pertaining
to the matrix increment spacings for MTF and ground exposure, the uniqueness of
each parameter, the precision of the judgments, and the ability of interpreters
to rank mission material with GEMS.
2.5.1 Parameter Scaling Findings
The study results established that equal, visually discrimin-
able steps of MTF and exposure aro obtained for the ranges of the parameters
tested if the increments of,MTF are spaced at equal logarithmic steps and the
increments of exposure shift are spaced at equal density steps. A 671 accuracy
of discrimination is obtainable when the MTF steps are established at 12.5%
ground resolution increment intervals and the exposure steps are established
at 0.06 density shift intervals.
The results imply that-on exposure shift interval of 0.06
density units is too fine tar camera system evaluation purposes since no changes
Approved For Release 2002/06/17 1 cIA-RDP78B04747A000700010025-7
 Approved For Release 2002/06117 CIA-RbP78B04747A000700010025-7
in resolution due to exposure were discernable for the entire simulated exposure
range of the GEMS.. For the GEMS matrix, it is suggested that the exposure
increments be established to correspond to exposure shifts.that produce 12.5%
changes in limiting ground resolution.
An MTF interval of 12.5% ground resolution steps is as fine
an increment spacing that can be discriminated with reasonable accuracy. It
is suggested that this interval be established as the increment spacing of the
GEMS matrix 'array. With the use of a split field viewer for making comparisons,
the accuracy of M.`F discrimination may be improved.
2.5.2 Other Studz, Findings .
The experiment concerning the ranking of mission material
with a GEMS matrix indicated that the subjects' judgments of the MTF and
exposure levels of the mission photographs were just as rz.liable as their
judgments when ranking one GEMS against another. In other words, the standard
deviations of their judgments for each parameter were statistically the same
in both experiments. The a, curacy of -the judgments could not: be determined
since no objective measures were made of the mission photographs used for t:he
study.
A very important finding of the study was t1-_aL scene content
seemed to have no effect on-the subjects' ability to judge MTF or exposure.
This finding strongly suggests that the parameter of scene content can be con.-
fined to few scenes of various type imagery.
2. 6 GEMS VIEWER STUDY
A concept design study was performed to define features most
desi.;-r.ible for a GEMS Viewer. The established instrument design concept will
13
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 : CIA-RDP78B04747A000700010025-7
enable rapid image quality assessment of roll film aerial photography with GEMS.
The instrument was designed to provide rapid assessment capabilities by con-
sideration of the following features:
1.
Means for easy storage and quick retrieval of a library of known-
quality images. The retrieval system will enable an operator to
select a GEMS having known quality values.
2.
Means for rapid screening of original negative roll film.
The
instrumedt will include a motorized roll film transport, designed
to insure safety of roll film, and a variable intensity light
table.
3. Means for viewing the roll film images and the reference images
(GEMS), simultaneously.
4. Means for manipulation of the two images to permit conven:Lent
positioning for optimum viewing.
A dual split-fi_ci' presentation was chosen as the optimum device for
viewing two photographs simultaneously. The operator will see a spl= field,
each half containing an image to be compared.
This viewer must not degrade the images and yet must contain a variety
of adjustments to allow the' operator to magnify and crient the images to suit
his individual preference. It was necessary to design an optical system
specially for the GEMS Viewer in order to accomrodate the unique optical paths,
maintain the resolution of the images, and provide the magnifications and :i_mage
motions required to permit rapid comparison of photo graplhy.
It is also important that the comparisons between the unknown images
recorded on roll film and the known images recorded on GEMS be done. in rapid
Approved For Release 2002/06/171a.CIA-RDP78B04747A000700010025-7
 Approved For Release 2002/06717: CIA-RDP78B04747A000700010025-7
succession, with a minimum of operator distraction. This requires a fast storage
and retrieval system with a simplified' call up. A commercial address type
system was chosen for this applicaation. Mounting four GEMS. per address (slide)
increases the library storage capacity to 1024 GEMS.
Images recorded on the roll film must be located and positioned in
the field of the microscope and oriented to the GEMS images. Because this may
be irreplaceable material, attention has been given to the film transport
system to ensure the scffety of the film. Constant tension. is required in order
to maintain focus and prevent inching. A variable speed film transport is
required to minimize search and positioning time. -
The instrument is to be mounted on a hard wood table with the micro-
scope optical axis at a 15 degree downward. tilt. Pertinent aspects of human
engineering have been carefully considered to minimize operator fatigue and
maximize convenience and speed of operation..
The operation of this instrument will be simple even though it: con-
tains a variety of features. Many of these features,. although important, are
adju.strnents used only on. a "once per'r.ol_l" basis (or,even less frequently).
In addition, some operations are combined (e.g. joy sticks to control both X.
and Y mo-.; ,ments, simultaneously) . Further, all the controls have been care.--
fully located to facili.tat e manipulation at minimum operator effort. No
problems are anticipated in, the operation of this equipment over long periods
of time. -
While this i_rstrument has been designed to permit mission diagnosis
by comparison of photography and GELS, it is clear that: the instrumert.cen be
easily applied to several. (Aber aspects of the overall photo- i_nter.pretati.on
Approved For. Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 CIA-RDP78BO4747A000700010025-7
process. For instance, with little or no modification, the instrument can be
used for training of. new personnel, for assistance in target identification,.
and. for change detection. Other potential. applications may become apparent
later.
2.7 PSEUDO GEMS VIEWER STUDY
A Pseudo GEMS Viewer optical breadboard was constructed. The bread-
board was used to test the concept of manipulating the photograph scene parame
tors of HUI?, exposure, and contrast with optical and electrical mechanisms.
Such a device, if it produced realistic simulations, would eliminate the need
of a large matrix library of GEMS.
The study showed that reasonabl2.matches could be obtained for the
scenes employed, but theoretical, considerations imply that: the technique is
inadequate for the simulation of the appropriate mission material without:
further refinement. It was established under the Refinement of Techniques
task that the physics of a hazed aerial. scene is actually the reduction of the
ground imagery signal level by a superimposed, non--image forming, d.c. light:
level in the exposure domain..
To achieve an accurate haze. simulation in the Pseudo GEMS Viewer with
only the use of non-ima e forming light and a transparency, the density dis-
tribution of the viewer traiispa, .ency must be made linaarly proportional co '_he
original ground exposure distribution. In the above arrangement, a linear_
relationship between film density and ground exposure only exists for. a posii.ive
transparency whose cascaded gamma is unity. Since the pseudo concept must
employ a negative transparency with a ga~rma of approximately 2.2, the resulting
haze simulations will not be valid.
16
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
The pseudo breadboard experiment did not identify this flaw in the
concept, because positive transparencies with a cascaded gamma of unity were.
employed. however, by-inspecting the degraded photograph supplied in the
Pseudo GEMS experiment report', it can be seen that the low and high density
images were not identically matched. This mismatch at the density extremes
occurred because the toe and shoulder areas of the film characteristic curve
departed from unity gamma, If gamma'2.2 negative transparencies had been.
employed in the breadboard experiment, considerably more of this type of dis-
similarity would have resulted.
Exposure shifts also cannot be accurately simulated by the experi-
mental pseudo approach, because of the negative transparency's non-linear'ities.
The GEMS concept eliminates the non-linearity problem by e mploying a unity
gamma, positive, master transparency in the simulation:.process.
Approved For Release 2002/06/17' : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
PROGRAM CONCLUSIONS.
3.1 TECHNOLOGICAL STATUS
The efforts of the GYMS Study program can be termed quite successful.
All of the major objectives of the program have been achieved. Studies per-
taining to the development of simulation techniques have led o the establish-
ment of means for accurately controlling the production of realistic photo-
graphic simulations.
The problems associated with generating GEMS at the same scale factor
as mission material were defined, and a procedure has been specified for elimi-
nating all apparent simulation dif.ficu1:ties. The lcey factor that allows the
generation of realistic simulations is the possession of .a positive master
transparency whose image density distribution is-linearly proportional to the
ground. exposure distribution.
In the sirnilation process, such a transparency .permits the modifi-
cation of image spread functions and. the introduction of ground exposure shifts
and varying degrees of atmospheric haze in identically the same manner as what
normally occurs in real life, except for the effect of atmospheric haze back-
scatter into shadow areas. Fortunately, subjective estimates of contrast are
not dependent upon the phenomena of back-scatter.
The realism of the GEMS simulations was substantiated in the psycho-
physical study where it was demonstrated that photo-interpreters could sub-
ject.iuively estimate the image quality of mission material wilt the same degree
of reliability as the ranking of one GEMS against another. The psychophysical
18
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
 Approved Fo- elease 2002/.0.6/17: CIA-RDP7.8BO4 47A 007000100 5-7
study also es abl:i.she tie matrix >ncremen spat] ngs tor ~e parame er s of MTF
and exposure as well as the accu..racy of the judgments of each of these parame-?
3.2 USEFULNESS OF GEMS
It is. essential to establish that the general concept'of GEMS is use-
ful. The usefulness of GEMS can be determined by examining the accuracy with
which the parameters can be judged. The psychophysical study results demon-
strate. that a photo-interpreter can judge both a MTF increment spacing of 12.5%
ground resolution steps and an exposure increment spacing of 0.06 density unit
steps with a 67% confidence of proper discrimination. Their ability to dis-
criminate levels of exposure is better than is necessary. Although their
discrimination of MiF levels may be slightly less sensitive than desired,
reasonable estimates of NTF levels are obtained.
Serious consideration should be given to the point that the GEMS
estimates of MIT are almost as accurate as the MTF objective determinations of
mission material, because of the variable factors associated with the typical
targets used in the measurement process. Under certain conditions, where the
quality of imagery is varying over the format and no objective measurement
targets exist in the areas of interest, the MTF estimates obtained with GEMS
can be much more accurate than any infe.rrcd evaluations. Image quality varia-
tions occur quite frequently duo to variable haze conditions or variable cloud
coverage exposure conditions.
The attributes of GEMS are another factor to be considered in deter.-
minim; the usefulness of GEMS. The major attributes of GEMS are that they can
supply estimates of system performance quite rapidly and that these estimates
are not dependent: upon either specific type targets or complicated instrumen-
1S
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
tation that requires lengthy operator training time. It probably would require
a minimum of 4'to 8 hours t:oobjectively evaluate, without an on-line computer
to the instrumentation, the parameters of a single scene area that could be
estimated with a GEMS matrix in a matter of a few minutes.
In summary, the psychophysical study demonstrated that GEMS can
provide a useful service as a rapid evaluation tool.
3.3 PROGRAM RECOMMENDATIONS
At the last program review conference with the customer, 5X1A
stated that it would objectively evaluate both the GEMS and Pseudo GEMS con-
cepts; and to recommend on the basis of the technical soundness of eacl:
approach, the concept that would best serve the customer in his systems' evalua-
tion work. The findings of the various. study tasks definitely establish, with-
out any doubt, that the GEMS concept is the most valid approach for acquiring
accurate diagnostic estimates of system performance at this ti1ma,
When considering the facts, the simulation technology has been
developed, the psychophysical parameter. scaling has been -determined, and the
appropriate. system parameters have been defined, implementing a GEMS matri.,
as an evaluation tool can be accomplished. in a fairly routine fashion. ii-
amounts to generating the GEMS library material and fabricating a GEMS Viewer
which has already been designed. It is our reconcaendation that the GEMS con-
cept be implemented.
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
 Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7
Approved For Release 2002/06/17 : CIA-RDP78BO4747A000700010025-7