SPECIAL DIGITAL IMAGE MANIPULATION PROJECT

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B05703A000100070008-0
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 2
• Document Creation Date: December 28, 2016
• Document Release Date: January 20, 2004
• Sequence Number: 8
• Publication Date: June 29, 1970
• Content Type: MF

5118.331CT
-
Approved ForReose 2004/02/11 :.CIA-RDP78B05703AM100070008-0
11,46,
SECRET
OPTC/TSEG D/ATB-127/7n.
Pj June 1970
LD/&&T/eld Of tee
Executive Director, VIC
Chief, Advanced Technology lira RID/TSOG
Chief, MSvearch &Engineering DI ion, TWO
Chief Technieal ::ervicee it Support Group, VPIC
Special Digital ImaEc Manipulation Project
,or investigating Dtitsi I
support or a special projec
Office, this research foal
of app1yin4 Digita image Mani
d
tion (
DIM) technique*,
OWS&Thrield
uoett to demonstrate the fss,ib1tty
tion technir,ume to a certain class
of operatiooal problems. This memoranduadiscuzsee thic- exercise
bicIvas performed during the period 8-14 April 19TO. Detail of
the experiments are reported as attachments.
2. The data for this study originated from two entirely different
low re4olution, terrestrial collection systems. In one CAA! (attacheent
#I) the data was collected through en optical system and recorded
directly mato fi;e4 whereas lo the other case (attachment 0) a T.V.
system was Ivied to record the data on video tape. The data supplied
to XPIC for input to the um system, was second generation and was
entirely on photographic film. In the first case the original nege.
tive (tri.x pea) was contact printed as a direct positive (4427) end
in the second case, the T.V. data was bulged on a cathoie ray tube and
photographed (film unknolon). no further information WS supplied.
3. Ihe results of the study.indicate that for these collection
systems, DIM may offer sone *existence with mensuration an photointer-
pretation problems. With each image studied, edges 'were better
defined end the overall noise level vas reduced. In practice, tie
better definition Nould make it easier to deterelne the dimensions oh.
the objects of ietsre-it, and in the preenee of less noise, provi44,
additanal insight concerning their shape.
DECLASS REVIEW by NGA/DOD
?CRETA
Approved For Release 2004/0 1 : CIA-rcIDP
GROUP 1
Excluded from automatic
downgrading anti
oz,:q o .04 I I
00070008-0
,(---08= I
25
4
Approved For ReIse 2004/02/11: CIA-RDP781305703A0100070008-0
SECRET
SUBJECT: Special Die, Project
4. Ttere mire ettieant 1taititLns encolntered in till! taxer-
ci-e, i.e., insufficient date ecernn; the input, reaesrch oriented
hardwire and software, and an imeo severely glegraded by photographic
grain structure. A. a cocieuence, the result) reported here represent
the state of the technology as applied only under poor condition% ane
o not reprirent what woold he obtains with an operational ystem
designed to satitfy operational needs. For thi problem, the applica-
tiou cJf image eanlpulaton technitmcs WAX analogous ta * mathematical
exerci);e, here one ia given an approUmate solution (the image), an
then rered to formulete the problem (Oiagnose the probleem in the
image) RAC: calculate a new *weer (manipulate the image). ulth cali-
brated input data, better reaultg c1i have been obtained.
5. DIM cannot effectively ctin *a an tgo1ete4 interpretation
techniue, tut must Oa treated as an integral part or the overall
collecton eyntem. Therefore, the routine applicetion of DM tech-
niusts. to any operational problem -will neceagitate the re-c4aminatiou
an celibration of the collection system an tte development ..7.1!
modified DIM techai.4oe e_jiipment. Pr thee.;* particular protaem,
thls say reTcluire the modification of the collection system, the uee
eirw
record tnt. materials and proceasing, and the expansion of tne
DIM systea to analyte video as t,ell 45 photographic data.
o. DeRpite the probleme and the limitattils encouetered h,s
t anis demonstrated the feasibility of applying DIM tech U.6
perational protasis, ha,i permitted the tranelation of theoreti-
cal onliderstion-.5 into prattical terms, aaJ haa helped identify
prolem areas. These resut, have impact on the DIM program in that
they are generically related to MC probiase, and provide siznificant
perspective with respect to the errent state of applied technology.
/ATB/ITL
Diatributiln:
Orig. & 1 - Addreasee
1 - Exec. Dir/EPIC
1 - MPIC/TESG
2 - DTIC/ISSG/RED/ATE
SECRET]
Approved For Release 2004/02/11 : CIA-RDP78605703A000100070008-0