RESEARCH AND DEVELOPEMENT PROJECT APPROVAL REQUEST

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B04747A003100010063-8
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 6
• Document Creation Date: December 28, 2016
• Document Release Date: February 9, 2000
• Sequence Number: 63
• Content Type: REPORT

Approved For Release 2000/G lOF'eW-= 8B04747A003100010063-8
Research and Development
Project Approval Request
1. Identification
This project is the follow-on to the Drell Committee, and will carry
out development of procedures for image quality evaluation on aerial photo-
graphy. Its internal designation is, "Image Quality.Evaluat on Program."
The project will come under the Tec lopment program of the P&DS,
NPIC, at a total estimated cost of It will involve contractual 25X1A
obligations to five industrial firms, and to the National Bureau of Standards.
It was not originally included in NPIC's 1964 financial plans, and funds
are not available without deferring other projects scheduled for this fiscal
year.
2. Objectives
The program will develop and validate a procedure(s) for both technical
and routine analysis for determining the quality of aerial photography, applying
microdensitometric edge-trace (or any other analytical technique of equal
or improved precision which ma or may not involve edges) and comparative
25X1A photography, designated as The method and procedures must be suitable
for routine production, using automated techniques, but expediency must not
be substituted for accuracy.
3. Background 25X1 A
The report from the llllllllllllllllllllCommittee established the need for examination
of operational photography through the technique of edge-trace and _ to
assess quality (through derived knowledge of the system transfer function),
and provide guidance for analysis of system errors. Following this report,
a memorandum was issued by Mr. Lundahl outlining a plan for implementation
of this by NPIC. Subsequently, a committee convened at the instigation of
NRO end published a.memorandum outlining four (4) basic tasks, of which
the Image Quality Evaluation was one. A work statement was issued, outlining
the aims of the study, and suggesting a committee consisting of industrial
and governmental organizations be formed to do the necessary work. This
project approval request is based on that work statement. All the insti-
tutions listed in the work statement have been contacted, and proposals
received in reply. 25X1A
The program will be a..direct extension of the Committee recom-
mendations, with an allowance made for additional considerations in pursuit
of the same goal, The efforts of the group will be focused at`.NPIC..
CONFIDENTIAL
25X1A
GROUP .1
E.c l.d :J. f~.m -...iic
dew nq.odinnd d.closeli-i-
Approved For Release 2000/05/10 : CIA-RDP78BO4747AO03100010063-8
 Approved For Release 2000/05 Q~ 4e AE P7%B04747A003100010063-8
Some investigation and development are already underway by the Plans
and Development Staff, and these will supplement the work to be carried out
by the committee. The group is made up of representatives of industrial
and governmental organizations actively engaged in image quality evaluation
techniques and data reduction. A nine to twelve month effort by this group
of expert talent will define an experimental program and send out tasks
(contracts as required) to competent laboratories (NPIC, SPPL, BuStandards,
FTD, etc.) and specific contractor laboratories to complete the experimental
program outlined.
The group plans to meet regularly to plan the technical developments
and review the results. The experimental program will be periodically
revised to reject as quickly as possible those methods of limited promise,
and to arrive expeditiously at standard procedures.
4. Technical Specifications
The group membership is as follows:
Chairman:
Vice-Chairman:
Technical Advisor:
Members:
STATOTHR
25X1A
no STATINTL !
The individual members of this group have all submitted proposals for
undertaking various aspects of the work. In general, there is sufficient
overlap of those important areas where some duplication is mandatory for
scientific accuracy. All of the relevant areas of the problem have been
covered. The group will initially meet at; ;NPIC, learn of each other's
program, establish a common program philosophy, and establish working
procedures. This initial meeting will be based on the following pre-
liminary outline:
a. Conduct a review of work carried out to date, both pub-
lished and unpublished.
b. Survey and analyze the characteristics of microdensitometrics
available to group members and organizations, tasked by the
group as well as other commercially available instruments
suitable for edge measurements.
CONFIDENTIAL
, Perkin-Elmer Corp.
Itek Corporation
Eastman Kodak Company (Representative
as yet unassigned)
Approved For Release 2000/05/10 : CIA-RDP78BO4747AO03100010063-8
 Approved For Release 2000/D51W!`CTA=kI678B04747A003100010063-8
c. Review other quantitative and qualitative measures of image
quality to be investigated*
d, Recommend the laboratory and/or organization and define the
work to be accomplished there in support of the objectives.
Following this initial meeting, the program will be carried out at uentl the contractors' fact-ies. rland program evadluat1ion. Theyremainder
re-orientation, outline,
quired for review, -
carried out according to the following of the program will be car
approximately.
Investigation and Experiment.
a. Specify and prepare sample edges, comparators, or other
materials for circulation and/or assignment to specific
epared by
organizations. These materials to be pr a cali-
brated image forming system with known transfer function.
Emphasis to be placed on 4404 film, but a.-second film
(4400 should be included for comparison.
from
b. Procedures, measurements and data reduction resulting
the supplied materials of each laboratory and/or organiza-
tion is to be provided all the other laboratories and/or
organizations.
c. Group review so as to make modcations ofetechniques and.
repeat the experiment. This procedure
a concensus is achieved on proper procedure
d. Evaluation of results must at all times be guided by appro-
priate statistical measures of systematic and random errors.
e. Techniques should be recorded and rejected that are not
susceptible to application by trained technicians in a
routine way.
Evaluation:
the approved techniques to actual aerial photographs.
a. Apply
b. Analyze these results statistically to determine the accuracy
and reliability under different conditions.
c. Publish procedures that apply to collection systems and that
can be used by trained technicians.
CONFIDENTIAL
Approved For Release 2000/05/10 : CIA-RDP78BO4747AO03100010063-8
 Approved For Release 2000/057/OFICFA-RDP78BO4747AO03100010063-8
5, Contractor and Financial Arrangement
Various aspects of this program will be placed with the corm nobs.
listed below, for thet'.amounts shown. The progacam will cover. 9-
25X1A
Each of these firms has been chosen because of its prominance in
nhthe fieldl
STAT R cf Image Quality Evaluation. All but Committee, and therefore she fully
25X1A Eureau of Standards were on the Allllllllllllllllll cognizant of all the problems involved. This project constitutes arr.
investigation into ill defined and unexplored realms of knowledge for which
the work to be performed cannot be precisely defined. Therefore, it is
mercyosFd that this work be negotiated on a cost plus fixed fee basis.
program has been worked out in cooperation with NRO and is
known (and approved by) the DD/S&T. All results will naturally be made
available to these components.
Security
erate this group on an unclassified basis, except
o
d t
p
o
It is planne
for one member of each institution who will be fully cleared for access
to all relevent data and material at NPIC. SC-l, Confidential, should
suffice for all contractual arrangements.
CONFIDENTIAL
Approved For Release 2000/05/10 : CIA-RDP78BO4747AO03100010063-8
 Approved For Rel, se 2000/6 B IA-RDP78BO4747AO03100010063-8
A;ril 1963
went
Ezemtive Secretary,
tble
Of itrureet
1*4 film capabilities In
esirble results as far the types of
xn, filtering concepts leading to
"is- tusd a vs.i bl.lity is r l , ovalustion should be mWW Web vOulu
csatract -rim. ena'ble to obtain
th tbase amidifi ticmx indicated in item 1. To
t ~ s no possibility of using c 1estfi film for efte enhancement.
Of zr i nts atte bie at Imp stile of the viewer ann.
Visibility for accentust# low contrast image edges. At
ace nt Ve er will duce, through the use of s .tii1
rre tt y haves certain f e tcies-
Bch s: dement up= the size
me, is imge edge cement.
se of de to nt is n x$ti L but x .ifi-
;orpoxat in wee IT the contract exten
se of o ticm.
~ T
me II
Ov dradinv and
de. 0a5SltiCaiiCn
Approved For Release 2000/05/10 : CIA-RDP78BO4747AO03100010063-8
 Approved For Rase 2000 1CIA-RDP78QD#747AO03100010063-8
rder t.t the it'd' sru4 uw :...._ _
rut s a thmt sub ueUt amlymis be mile of Sel&iti :1 ph tv 1hY
the r qu. to erect,
Led natre at , It is
(a. O.
t View
+
;berm ~in~ for Xe
r try as
f
er
qVroved, The C=txact extengi , re
hit to
six maths f_ -te o
uent to this 1riod. at a total cost9, 186.
25X1A
t*vclopment , POS
El
Approved For Release 2000/05/10 : CIA-RDP78BO4747AO03100010063-8