FEASIBILITY RESEARCH ON A SYSTEM TO PROVIDE HIGH RESOLUTION PHOTOGRAPHY OVER DENIED AREAS

• Document Type: SPECIALCOLLECTION
• Collection: Animal Partners
• Document Number (FOIA) /ESDN (CREST): 06527327
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 198
• Document Creation Date: September 6, 2019
• Document Release Date: September 12, 2019
• Publication Date: April 1, 1978

Approved for Release: 2019/07/31 C06527327. A] 6, �
cc(--
To titor__,
NOFORN
6.2(d)
3.3(h)(2)
Feasibility Research on a System to Provide High
Resolution Photography Over Denied Areas
ret
Handle via
TALENT-KEYHOLE Channels
RD-10-77
April 1978
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Copy N1.1
2
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se �UFF
OFORN
Feasibility Research on a System to Provide High
Resolution Photography Over Denied Areas
A Research Study
By
Charles N. Adkins
Operations Technology Division
Office of Research and Development
ED.
Chief, Oper ns Technology Division, ORD
Handle via
TALENT-KEYHOLE Channels
TCS 35974-77
III
ecret
April 1978
Date
E2 IMPDET
CL BY
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Secre
�FORN
TABLE OF CONTENTS
Page
Abstract
1
Acknowledgements
1
Part1�Introduction . ....... .................... ............ ..............................
3
Part 2�Target Description
5
Part 3�Camera Development
21
Part 4�Two Example Targets in the Washington, D.C. Area
29
Part 5�Bird Behavior and Statistics
51
Part 6�An Example Mission Scenario
67
Part 7�Summary and Conclusions
75
APPENDICES
Page
A. Evaluation of Photographic Coverage
79
1. Summary of Results ................ .................... ........
79
2. Determination of Exploitation Suitability
81
3. Image Quality Analysis & Film Selection
89
B. Camera Detail and Special Devices
131
C. Care and Handling of Homing Pigeons
143
D. A Brief History of the Pigeon ............ .......... ..............
151
References............. ......... ...................... .............. ..........
155
FIGURES
Page
2445 Color of Museum Park .............. .............. ........ Facing Page 1
1. Satellite Photo of Sudomekh Ship Yard 7
2. Satellite Photo of Admiralty Ship Yard 9
Handle via
TALENT-KEYHOLE Channels
TCS 35974-77 Isp-Setr
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Secr
ORN
Page
3. Satellite Photo of Baltic Ship Yard 11
4. Street Map of Leningrad ............ ......... ............... ....... ......... 15
9.
MCW-24 Camera 23
10.
Balloon Picture
Using 3400 Film 24
11.
Camera..... ................:.. ............... ......... 26
12.
Bird with
Camera 27
13.
Four Flight Paths Over Andrews Air Force Base 30
14.
Mobile Home Complex on Andrews Air Force Base 31
15.
Military Trucks on Andrews Air Force Base 32
16.
Incinerator Plant on Andrews Air Force Base 33
17.
Military Aircraft on Andrews Air Force Base .............. 34
18.
Satellite Photo of Andrews Air Force Base 37
19.
Map of the Washington Navy Yard 39
20.
Corner with People Walking to Work 40
21.
Alley Way 41
22.
Roof Top with Air-Conditioner 42
23.
Parking Lot 43
24.
Oblique Shot of Museum Park 44
25.
Navy Yard Main Gate .................... ......... ............. 45
26.
Old Naval Gun Factory 46
27.
2445 Color of Museum Park 47
28.
Oblique Photograph Over the Navy Yard ...................... 48
29.
Satellite Photo of the Washington Navy Yard 49
30.
Single-Toss Experiment 54
31.
Histogram of 84 Paired Flights 55
32.
Chance of Overflight vs. Subtended Target Angle 56
33.
Bird With Camera Returning From the Navy Yard 57
34.
Exterior View of the Attic Loft ................ ........................ 61
35.
Interior View of the Attic Loft 62
36.
Reinforcement Training to Drop Hole ............ 62
37.
Bird Leaving Air-Conditioning Shell ....... .................... 63
38.
Widowhood Loft with Air-Conditioning Shell 65
39.
Hours the Sun is Above a Given Elevation for January-June
in Leningrad 67
Handle via
TALENT-KEYHOLE Channels ' vi
TCS 35974-77 ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Sec RUFF
FORN
Page
40. Hours the Sun is Above a Given Elevation for July-
December in Leningrad 68
41. Pigeon In Flight After Clandestine Release 72
42. Miniature Pan Camera 76
43. Pan Picture with 180� Field of View 77
APPENDIX A FIGURES
Al
A2
A3
A4
A5
A6
A7
A8
A9
A10-17
A18-20
A21-22
A23-27
A28
A29
Handle via
TALENT-KEYHOLE
TCS 35974-77
Page
Histogram of NIIRS Ratings of Tacana and KH-8
Imagery 98
Histogram of Cumulative Frequency of Error in
Measurements of 53 Dimensions (.25 Feet-37.5 Feet) 99
Graph of Cumulative Percentage Error for Measure-
ments of 53 Dimensions 99
Histogram of Cumulative Frequency of Error for 22
Dimensions of 2 Feet or Less 100
Graph of Cumulative Percentage Error of 2 Feet or
Less 101
Histogram of Cumulative Frequency of Error for 14
Dimensions of 1 Foot or Less 102
Graph of Cumulative Percentage Error for Measure-
ments of 1 Foot or Less 103
Graph of Predicted Target Brightness/Exposure
Range and Film Estimates 104
Graph of the Camera/Film Performance Evaluation
and GRD Estimates 105
Examples of Vertical and Oblique Photographs
NIIRS Rated by NPIC PI's - 106-113
High Altitude Acquisitions 114-116
Adjacent Frames Demonstrating the Effects of
Unpredictable Platform Motion (Frame-to-Frame) 117-118
Successive Frame Coverage of the Navy Yard
Target Area 119-123
Example of Navy Yard Target Coverage by the Model
I (MCS-24) 7.5MM x 12.5MM Format Camera, Not
Mensurated 124
Ground Views of TALOS Missiles 125
Channels
VII
Top S ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Secret
�RN
A30 KH-8 Coverage of the Navy Yard Museum Display
Page
(Target) Area
126
A31 Color Photograph Taken by Tacana Over the Wash-
ington Navy Yard, 7/8/77
127
A32(a-d) IDIMS Effect on Smeared Imagery
128
A33 Camera�Film Analysis Test Target
129
APPENDIX B FIGURES
B1
B2
B3
B4
B5
B6
B7
B8
B9
Page
131
Camera Mechanism Assembly 132
Timer Box Schematic .............. ................... ............... 134
Timer Box Controls ....... ......... ................... ............... ......... 135
Camera Electronic Timer ....... .................. ........... 136
Camera Harness 137
Bird Responding in Skinner Box ................... ......... 139
Bird Receiving Food Reward 140
Infrared Strobe and Timer 141
MCW-24 Camera Mechanism Assembly
TABLES
Page
1. Key Areas in Sudomekh Ship Yard 13
2. Key Areas in Admiralty Ship Yard 13
3. Key Areas in Baltic Ship Yard 13
4. MCW-24 Characteristics 22
5. Camera Data 25
6. Loss Data for Kit 1F 52
7. Flight Performance for Kit 1F 53
8. Flight Performance for Kit 2F 53
9. Disposition of 118 Birds 58
10. Relocation Experiments with 118 Birds 58
11. Origin Statistics on 101 Birds ................... ................... 59
12. Destination Statistics on 101 Birds ..... ......................... 59
13. First Relocation Statistics 60
14. Flight Performance for Relocation Kits 2H and 3B 60
Handle via
TALENT-KEYHOLE Channels
TCS 35974-77 Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Secr FF
�FORN
APPENDIX A TABLES
Page
Al Measurements made from Tacana Navy Yard Willard Park
-Target- Imagery 85
A2 Tacana System Parameters 90
AS GRD Estimates of System Performance over Andrews AFB 92
A4 Materials Discrimination Table of the Special Color Analy-
sis Technique (SCAT) 95
A5 Characteristics of Black and White and Color Films
Considered for Tacana Applications 96
A6 Weighted Scoring Table for Film Selection 96
A7 Film Performance Scores 97
Handle via
TALENT-KEYHOLE Channels
TCS 35974-77
ix
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Secr FF
ORN
2445 Color of Museum Park SE ET
Handle via
TALENT-KEYHOLE Channels
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se
RUFF
FORN
FEASIBILITY RESEARCH ON A SYSTEM
TO PROVIDE HIGH RESOLUTION
PHOTOGRAPHY OVER DENIED AREAS
ABSTRACT
_CIS} Research was conducted on a system (Tacana) which may
provide high-resolution photography over denied areas by the use of
homing pigeons. The pigeons receive no specialized training; they
fly from the release point across the target to the home loft (e.g.,
etc.). An adjustable timer in the camera starts
the photographic coverage over the target and lasts for about four
minutes (150-220 pictures). Each picture covers an area on the
ground about 90 feet square. The research was directed toward
photographic coverage of denied areas where direct overflight is
possible. Coverage of the Soviet Nuclear A-class submarine work at
the shipyards in Leningrad was considered as an example high-
priority target. Two simulated targets (Andrews AFB and the
Washington Navy Yard) were chosen to provide data on system
performance, including in-flight photography. The photographic
analysis was performed by NPIC and compared with overhead
satellite photography and the specific intelligence requirement in the
Leningrad area. It is concluded that there is a good probability that
homing pigeons can be used to satisfy the high-resolution photo-
graphic requirement in the Leningrad area. A 16mm silent film is
available showing various aspects of this program. High resolution
prints of the Avian Photography are available under separate cover
as a supplement to this report.
ACKNOWLEDGMENTS
4�1 The National Photographic Interpretation Center (NPIC)
provided all photographic development and analysis. This was
coordinated through (ISB/ APSD/TSG) who also
directed the photographic experiments and recommended the type
film and film development best suited for this particular system.
4..P31 Acknowledgment is also extended to the Deputy Director
for Intelligence (DDI), whose offices (OIA and OWI) provided data
associated with the Agency's high-resolution photographic require-
Handle via
TALENT-KEYHOLE Channels
TCS 35974-77
1
Top ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top et RUFF
ment in the Leningrad area. The DDI was also instrumental in
arranging the temporary reassignment of (an
avid homing pigeon enthusiast) from the Office of Strategic
Research to the Office of Research and Development.
was responsible for the selection of loft keepers in Oregon, Alaska,
and Virginia and directed the relocation experiments. He was also
responsible for collecting the photographic data over the two
simulated targets (Andrews AFB and the Washington Navy Yard)
which included all aspects of operation, maintenance, and field
repair of the avian cameras.
TCS 35974-77
2
Top ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top S RUFF
PART 1
INTRODUCTION
_LISt-T he idea of using birds for emplacement and photographic
coverage has been explored by the Agency for several years. These
studies invariably required some form of specialized training which
required the birds to respond in a desired, predictable way to a
specific recognized object. The purpose of this research is to
investigate the collection of high-resolution photography by the use
of homing pigeons which receive no special training other than
learning to carry a small avian camera. The pigeons fly from the
release point across the target to the home loft
An adjustable timer in
the camera starts the photographic coverage over the target and lasts
for about four minutes (150-220 pictures). Each picture covers an
area of about 90 feet square from an altitude of 100 feet and has a
resolution on the order of one inch (see Appendix A: Evaluation of
Photographic Coverage).
1,..P3") Though this research applies to any target where direct
overflight is possible, photographic coverage of the Soviet Nuclear
A-class submarine work at the shipyards in Leningrad is considered
as an illustrative example. Part 2 of this report discusses the high-
resolution photographic requirement in the Leningrad area and
shows the possible launch points, target sites, and loft locations
JcWr With the exception of the camera development, this
research was conducted in the time frame between September 1976
and July 1977. The program was designed to answer critical
questions associated with the relocation of homing pigeons, their
ability to collect photography over example targets, and the behavior
and statistics of their performance. Part 3 discusses the camera
development, and Part 4 describes two example targets in the
Washington, D.C. area (Andrews AFB and the Washington Navy
Yard) used to provide statistical data and in-flight photography. Part
5 addresses bird behavior and statistics for the relocation and
example target phases of the program. An example Scenario and
Mission associated problems are discussed in Part 6. Part 7 presents
the Summary and Conclusions and discusses certain areas in which
further research may be needed.
TCS 35974-77
13,1r7op cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
2ST The analysis of the photography collected over the
example targets was performed by NPIC and compared with
overhead satellite photography and the specific intelligence require-
ment in the Leningrad area. This analysis is presented in Appendix
A. It is felt that the overall results to date firmly establish the
feasibility of using homing pigeons to collect the desired high-
resolution photography in the Leningrad area. The total contractual
cost for this research was $78,000.
TCS 35974-77 T Secret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se RUFF
PART 2
TARGET DESCRIPTION
TCS 85974-77
;S,Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Sec FF
1
TCS 35974-77
6
cret
Approved for Release: 2019/07/31 C06527327
OS IMO MS MI 811 11"1 11�11 `ml` dmi` alma all SS NB MI OS
Approved for Release: 2019/07/31 006527327
2,2-PL6ST Sal
dr-lei 4 aS doi
Approved for Release: 2019/07/31 006527327
�
OS OS MI NO OOP IMO I'm'a dim� 1.1." 1111111 111.�1 1"1. OS OM ea Oil MR le
Approved for Release: 2019/07/31 C06527327'
21-PL6ST SO1
Approved for Release: 2019/07/31 006527327
Oh OS OS SO MO 111111 " OS SS IMO SO
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
13
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se UFF
1
TCS 35974-77
14
Top -cref
Approved for Release: 2019/07/31 C06527327
amo as as He min AM& AM& Old aloft Mini
mum.. Approved for Release: 2019/07/31 CO6527327 418 ell MI Ili
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 C06527327
Top S RUFF
11
17
Secret
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Sec RUFF
TCS 35974-77
Secret
Approved for Release: 2019/07/31 C06527327
ap app am an MO ark EMI dumb mem milk e, es us sis S.
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 C06527327
Top Se FF
TCS 35974-77
PART 3
CAMERA DEVELOPMENT
Background
ificant camera development started subsequent to the
tests conducted in the fall of 1975. During these tests,
pigeons flew with an MCW-22 camera which used 9mm film, a
Minox -Bull's Eye- lense, and weighed about 55 grams. The shutter
speed at this time was 1/200 to 1/400 of a second, and a very high
percentage of the pictures were excessively blurred. It was deter-
mined that the blurring was due to angular rates produced by the
six hertz flapping frequency of the bird. Calculations conducted in
late 1975 predicted a required shutter speed of 1/1200 to 1/2500 of
a second in order to obtain resolutions on the order of one inch per
100 feet of altitude. However, with the f/3.5 Minox lense, exposure
constraints precluded rates faster than 1/1400 of a second. There-
proaches were taken: first, research was initiated on a
f/2.7 lens to permit subsequent design of a very high
speed camera; and, second, a 1/1400-second system was designed
and constructed using the existing Minox lense and a 16mm film
format. This camera, called the MCW-24, weighed only 35 grams
and was first test flown in January 1976. Furthermore, this camera
contained two timing circuits (the MCW-22 used only one) which
not only turned the camera on at the predicted time-over-target, but
also turned the camera off at the end of the roll. This second timer
prevented excessive camera wear and increased the system reliability
to a great extent. The design of this camera also included a linear
motion compensation feature; the film velocity during the taking of
pictures exactly compensates for a forward ground velocity of about
36 mph at 100 feet altitude. This feature was verified by photo-
graphing bar charts fixed to the side of an automobile driven at
various speeds.
1,81 The MCW-24 was test-flown through the spring and
summer of 1976 for a total of about 30 flights. About 20 to 30
percent of the pictures taken during these tests showed a resolution
of one inch (or better) per hundred feet of altitude (100 feet is a
typical altitude), whereas about 30 percent also showed resolutions
of 11/2 to 2 inches, and about 40 to 50 percent were excessively
blurred due to flapping and high roll rates of the bird in turns. This
21
T ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top RUFF
was close t ed result and verified the prediction that the
high-speed camera would be required to obtain high
resolution a arge percentage of the time. Table 4 lists some data for
the MCW-24 camera. Here, resolution is defined by the measure-
ments from "bar charts- which were photographed from the birds
during in-flight experiments. Figure 9 shows the MCW-24 camera
with harness.
Current Effort
,(JerergrFF
with camera and harness. Other detail data are contained in
Appendix B.
,(gf At this time it is felt that the camera contains all
the features required to obtain the desired high-resolution photogra-
phy from a bird platform, and that sufficient research has been
conducted to adequately demonstrate feasibility.
FIGURE 11
Camera
TCS 35974-77
26
Top ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top et RUFF
TCS 35974-77
FIGURE 12. Bird With Camera
27
To cret
ET
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top RUFF
PART 4
TWO EXAMPLE TARGETS IN THE
WASHINGTON, D.C. AREA
Andrews Air Force Base
One of the first actions in the program was to obtain a
group (kit) of birds which could be used to collect data a ainst local
targets. Such a group, called kit IF, was purchased
located at , Virginia. These sir
all) had previously been trained only to the west (75 miles) of
The first training flight on this program was conducted on
4 October 1976, one mile to the southeast, toward Andrews Air
Force Base. There were two training flights at three miles and one
each at six and 12 miles. The sixth flight, on 15 October, was from
Andrews Air Force Base, 18 miles from the home loft at
On the next day, four birds were selected to carry
harnesses and weights the 18 miles home. The birds were then given
one day rest and, on 18 October, bird number 1F4 carried camera
number C-7 over Andrews Air Force Base; camera numbers C-9 and
less are MCW-24 models described in the previous section.
,(*) It had been very difficult to determine vanishing bearings
(final directions of departure) from the launch site due to the high
density of trees to the east of the Base. The film from camera C-7
was compared against satellite photos to determine the bird's
trajectory during the several minutes of photography. Figure 13
shows this first trajectory (labeled no. 1) was to the north of the
direct line home. Figure 14 is an example of the photography
(about 140 pictures) taken on this flight. For the next flight, the
launch point (no. 2 in Figure 13) was moved to the southwest and
the trajectory, though still north of the line home, was closer to the
runway and hangers. Figure 15 shows several military trucks parked
on the base: For the third trajectory, the launch point was moved
still further to the southwest, and the bird flew right up to the
runway before turning north. Figure 16 shows an incinerator plant
located in the southeast portion of the base.
-(-Wr At this point it was suspected that the birds might be
avoiding the runway because of the noise and aircraft traffic. One
particular bird, number 1F2, had been flying the 18 miles in about
33 minutes with a dummy weight and a 12 mph headwind. On 27
s nine in
TCS 35974-77
29
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 006527327
Top et RUFF
15 LGR

Li
� ,,
A
ILLAG
.
i
,
I> ''
..,,
e < , 1.8 i
�
ck
1
Path
-4
,
-,�
\
^
� .
..,.....,-...--
.
AN'WS
4,
a
t
"M. .
� ,

P4Moivf
-

IR
FO
�:7141
..s.).
...,-
�
I NGSTOQ
co co, IMANOR
-.a
1'6-L-7'i
ao
1-AeWad1W-i-AE,
0
kiiiiiiiiier
FIGURE 13. Four Flight Paths over Andrews Air Force Base
5 ET
November, this bird carried camera C-5, returning home in 34
minutes. The flight path is shown as trajectory no. 4 in Figure 13.
This flight was directly across both runways, resulting in the photo
of the aircraft shown in Figure 17.
fe By the end of October, three of the birds were lost (one
returned three weeks later) and eight more birds were added to kit
1F. Flights continued through December with a total of 11 camera
flights and 19 flights with dummy weights. As mentioned in the
previous section, analysis of the film showed that there was a higher
percentage of blurred photos (see Figure 17) than during the
previous summer tests due to a decrease in shutter speed from
1/1400 to 1/1000 of a second. Since the camera was to be
delivered shortly, these MCW-24 cameras were not modified. While
waiting for delivery of the camera, the birds were worked
to the east of
toward the Washington Navy Yard.
1
TCS 35974-77 Top
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 C06527327
Top RUFF
FIGURE 14. Mobile Home Complex on Andrews Air Force Base
TCS 35974-77
To)19c.p etK
ET
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se FF
FIGURE 16. Incinerator Plant on Andrews Air Force Base
TCS 35974-77
32
Top ret
Approved for Release: 2019/07/31 C06527327
$ ET
Approved for Release: 2019/07/31 C06527327
ret RUFF
TCS 35974-77
FIGURE 15. Military Trucks on Andrews Air Force Base
33
To ecret
RET
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE 17. Military Aircraft on Andrews Air Force Base
TCS 35974-77
34
Top ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
cret RUFF
1
,(*) On 21 January, bird 1F2 was released at Andrews with a
camera and a small avian DF transmitter (see Appendix
B) which totaled about 50 grams in weight. Just after release, the
bird was attacked by a hawk and managed to escape to the east. By
use of the DF equipment, the bird was located one mile distant on
top of a church, apparently unharmed. On the following morning,
the bird had not returned. With the temperature in the low teens,
the bird and camera were found in a packing crate behind a
shopping center some three miles northeast of the church. Though
this bird continued to fly well, it would never again perform
properly with either camera or weight.
,(gf Over Andrews Air Force Base, there were a total of 12
flights with the MCW-24 camera, 19 flights with dummy weights
and one attempted flight with camera number C-10.
Many of these flights were conducted during one of the coldest
winters on record in this area. There were no camera failures due to
low temperature. It is important to note that the body temperature
of a pigeon is 107�F. Generally, pigeons perform well in cold
weather and poor in extremely hot weather.
The 12 camera flights recorded most of the Andrews
complex with exception of the active runways. It is felt that the
birds avoided these because of the noise and aircraft traffic. The
primary difficulty in launch site selection was the inability to
observe the vanishing bearing among the numerous tall trees. This
required tedious work and time consuming delays in comparing the
flight film with satellite photos in order to determine how best to
adjust the launch point. Figure 18 is a satellite photo of Andrews
showing the surrounding terrain.
The Washington Navy Yard
...cs.) A group of young birds (Kit 2F) was purchased in late
December and trained to home during January 1977. By 17
February, these birds and Kit 1F were flying with weights from the
Washington Navy Yard to directly west of
the Yard. At this time, four relocated birds were also flying th
with weights. During the remainder of February and early
March, these birds were -single-tossed" (launched one at a time) to
collect data on individual performance. The goal was to overfly the
small museum park located between the Navy Yard museum and
the river. The park and four of the trial launch sites are shown in
Figure 19. The single-toss experiments did not work well. Almost
every bird circled for three to five minutes waiting for other birds
with which they could fly home. On 9 March, double-toss (launched
in pairs) experiments began with immediate improvement in results.
TCS 35974-77
35
Top� ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top UFF
In pairs, the birds immediately took up a heading in the direction
of home. The "Homing of Single Pigeons- is discussed in Reference
3, and their performance does improve with training. However, this
increases the training manhours by an order of magnitude. On 7
March, flights began with the camera C-10, resulting in
excellent pictures over the Navy Yard. The trajectory is shown in
Figure 19. Figures 20 through 23 show the quality of pictures
obtained on this flight; about 80 percent of the pictures were of this
quality. On 8 April, camera C-10 was damaged and sent back for
repair. Launch experiments continued from the 11th Street Bridge
site, and on 21 April, camera C-10 was flown again and lost. A DF
transmitter was carried with the camera, but no signal could be
found during an extensive search. The bird returned three weeks
later without the harness or camera.
c51) camera C-11 was received and flown on 29 April,
and camera C-12 was first flown on 6 May. By mid-May several
shutter failures had occurred which were eventually traced to too
deep an anodizing process which caused a structural weakening of
the shutter material. This problem was corrected and the cameras
were also modified so that new shutter assemblies could be installed
in the field should failures continue to occur. No further camera
failures occurred and tests continued through 22 June. Example
photography is shown in Figure 24 (an oblique of the museum
park), Figure 25 (the main gate), and Figure 26 (the old Naval Gun
Factory building). Analysis of this photography showed that a higher
percentage of the pictures (i.e., Figures 24 and 25) were blurred
from C-11 and C-12 than from C-10. At first it was thought that
the increased percentage of blurring was due to the high winds or
nervous birds. However, it was finally determined that the equip-
ment used to measure the shutter speed had malfunctioned, result-
ing in an actual shutter speed of only 1/1600 of a second instead of
the expected 1/2400 of a second. This problem was corrected, and
between 6 and 15 July, four additonal flights were made with
camera C-11 and six with camera C-12. These tests focused on
testing several "special films- supplied by NPIC. These Were:
1. Aero color negative 2445
2. FE 6526, a high-speed fine grain film
3. 1414, a high-resolution UTB film
4. SO-131, an infrared film
5. H&W Type 77 panchromatic
Figure 27 shows a color shot of one corner of the museum park. A
detailed analysis and evaluation of film and photography is con-
tained in Appendix A.
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top et ReF--
TCS 35974-77 TQwecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE 19. Map of the Washington Navy Yard
fe?5, Of these last ten flights, four were recorded as -hitting the
target.- It should be pointed out that the included angle of the
museum park, measured from the 11th Street Bridge launch site, is
10 degrees in azimuth. In Figure 4 the included angle of the
Sudomekh Yard as measured from the release point near the
is about 80 degrees in
azimuth. An acceptance angle of 80 degrees is equivalent to trying
to hit any part of the Washington Navy Yard from a release site on
the 11th Street Bridge. Of 84 paired flights from the 11th Street
Bridge, fewer than six missed the Navy Yard completely. However,
this was after the behavior of the birds was established and the
release site adjusted for maximum probability.
X On one flight, the camera was tilted to the side to obtain a
high percentage of oblique shots. If a pair of birds were flown with
cameras titled to the right and left, a large area to either side of the
flight path would be recorded. Figure 28 is an example of this kind
of oblique photography.
f,81 There were a total of 219 flights over the Navy Yard With
either cameras or weights. Seven flights were with the MCW-24
camera and 31 were with the new Camera. The
remaining 181 flights were with weights; 64 were with relocated
birds. During a series of 84 paired flights from the 11th Street
Bridge, 25 percent were visually recorded as hitting the target (the
museum park), and 54 percent missed the target by less than 75
yards. These statistics are discussed in more detail in the next
section. Figure 29 shows a satellite photo of the Navy Yard and
surrounding area.
TCS 35974-77
39
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Serf RUFF
TCS 35974-77
FIGURE 20. Corner with People Walking to Work
40
ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se RUFF
FIGURE 21. Alley Way
SE.QRET
TCS 35974-77
41
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
Top Secr FF
FIGURE 22. Roof Top with Air-Conditioner
42
Top ret
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
1
I
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top SerRUFF
1
FIGURE 23. Parking Lot
SE
TCS 35974-77
43
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
Top Se RUFF
FIGURE 24. Oblique Shot of Museum Park
44
Top ret
I
I
I
1
1
I
I
I
I
I
I
I
I
I
I
I
I
I
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se �UFF
FIGURE 25. Navy Yard Main Gate
TCS 35974-77
45
To cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Secr
FIGURE 26. Old Naval Gun Factory
TCS 35974-77
46
To ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se UF F
1
TCS 35974-77
FIGURE 27. 2445 Color of Museum Park SE ET
� 47
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 006527327
Top S RUFF
TCS 35974-77
FIGURE 28. Oblique Photograph Over the Navy Yard
48
To ecret
Approved for Release: 2019/07/31 006527327
3.3(h)(2)
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se F F
PART 5
BIRD BEHAVIOR AND STATISTICS
It is not enough to show that the camera can take
acceptable pictures over the target. Certainly of equal importance is
the capability of the bird to fly over the target and the frequency,
or expectation, of its performance. Data on performance were
collected almost entirely from local birds which were flown over
two example targets in the Washington, D.C. area.
,(3'1 Another very important problem is that of providing birds
of proven performance in the target area. This might be done in
several ways. One way is to test birds locally over example targets
and select those of good performance for relocation to the target
area. A second possibility is to take young birds to the target area so
that this is their primary home. These birds could then be relocated
in the Washington area and tested over example targets. Those of
acceptable performance could then be taken back to the target area
with a high degree of confidence that they would perform well at
their primary home. In either case, the issue of relocating birds
from their primary home to a distant secondary home is of great
importance. During this research, experiments were conducted with
132 homing pigeons which were relocated to lofts in Oregon,
Alaska, Missouri, and Virginia.
Performance Over Example Targets
� feie)' In early October 1976, the first group (Kit) of nine birds
was obtained located at
Virginia. This Kit, called 1F, was increased to 17 in number by the
end of October. These were all veteran flyers; three had won
diplomas in 300-mile races, and the rest had been trained to at least
75 miles to the west of The first training flight was
conducted on 4 October 1976; and the sixth flight, on 15 vetober,
was from the far side of Andrews AFB, 18 miles to the southeast of
On the next day, four of the nine birds were
selected to carry weights (the same shape and weight as the camera)
the 18 miles home. It was found very difficult for observers to
obtain vanishing bearings from the launch site due to the high
density of tall trees to the southeast of the base. On 18 October bird
number 1F4 was launched with a camera, resulting in trajectory
number one shown in Figures 13 and 18 by a comparison of avian
and satellite photography. It is apparent that this trajectory is
considerably north of the direct line home to It is
51
TCS 35974-77
Top S
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top et RUFF
speculated, as discussed in Reference 3, that this behavior might be
due to the effects of the previous training direction, which required
the birds to fly east, instead of northwest, to get home. In order to
compensate for this factor, the launch point was moved several
times to the southwest, finally resulting in trajectory number four
which crossed both active runways and photographed the military
aircraft shown in Figure 17. It is also speculated that the birds may
have avoided the runways due to the noise and aircraft traffic. For
this reason, and the difficulty in observing vanishing bearings,
attention was turned toward the Washington Navy Yard as a second
example target.
In late December 1973, a group of young birds (Kit 2F), 14
in all, was obtained and trained to fly during January 1977. Unlike
Kit 1F, these birds had never flown before. By 17 February both
kits were flying with weights from the Washington Navy Yard,
located directly east of Virginia. These
tests continued through mid-July 1977.
�IrS...) Of the 17 birds in Kit 1F, seven survived through July
1977, two were lost at the loft for unknown reasons, and eight were
lost in flight training. These eight were lost before the end of
November 1976. This implies that about half of the old birds could
not adapt to the regimen of being trained to carry weights and
"dropped out" early in the program. Table 6 shows the "Loss" data
for Kit 1F. Of the seven surviving birds, four showed exceptional
performance, and three were marginal. Table 7 shows the flight
performance of Kit 1F in terms of the number of times a weight or
camera was carried by each bird. These data include flights with
weights at short training distances which provided a gradual buildup
in carrying ability.
TABLE 6
LOSS DATA FOR KIT IF
BIRD
NO.
DATE
LOST
DISTANCE
MILES
NUMBER
FLIGHTS
3
1/19/77
0
*
7
10/05/76
3
2
9
10/29/76
18
9f
10
11/27/76
18
13
11
11/06/76
3
1
12
11/16/76
3
1
13
4/07/77
0
.
15
11/27/76
18
13
16
11/27/76
18
13
17
11/17/76
1
2
*Lost at loft, reason unknown.
f Four times to 18 miles, two with weights.
TCS 35974-77
52
lop cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top ret RUFF
TABLE 7
FLIGHT PERFORMANCE FOR KIT IF
BIRD
NO.
TOTAL
FLIGHTS
WEIGHTS
ANDREWS AFB
NAVY YARD
WEIGHTS
CAMERAS
WEIGHTS
CAMERAS
1
10
1
0
2
0
2
22
3
5
9
1
31
10
2
3
0
0
4
29
2
4
13
3
5*
21
4
1
4
0
6
25
2
0
12
3
8
30
2
0
20
15
9**
2
2
0
0
0
131
20
1
0
9
0
14*
8
0
0
3
0
*Placed on widowhood 3/3/77.
I Lost at loft, reason unknown.
**Lost fourth time from Andrews.
TABLE 8
FLIGHT PERFORMANCE FOR KIT 2F
NAVY YARD
BIRD NO.
TOTAL FLIGHTS
WEIGHTS
WEIGHTS
CAMERAS
2
15
11
0
10*
21
15
0
12
17
13
9
13
26
29
6
14
14
10
0
"Placed on Widowhood 3/3/77.
�(.81 Of the 14 young birds in Kit 2F, four were selected for
experiments associated with relocation; of the ten remaining, five
were lost on the very first release at the home loft. This is an
unusually high rate of loss and is most probably due to the late start
in their training. These birds were about 60- to 80-days old at the
time of their first release; typically, young birds are first released at
about 30 to 40 days of age, before they become too -wing strong"
and while they are less likely to fly off before learning where home
is. Normally, only a 10- to 20-percent loss is expected. Of these five
remaining birds, all survived. Table 8 shows the flight performance
of Kit 2F in terms of number of times a weight or camera was
carried, including several short training flights. As seen in the Table,
two of these five young birds performed exceptionally well and
carried a camera many times over the Navy Yard. Bird 2F10 was
placed on -Widowhood,- which is explained later.
"r The experiments at the Navy Yard were directed toward
finding a launch site from which the birds would fly over, or as
TCS 35974-77
53
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ret RUFF
close as possible to, the small museum park located between the
Navy Museum building and the river. This park contains many
naval artifacts, such as missiles, torpedos, small submarines, etc.
Ground truth data was measured on several of these items for use in
evaluating the avian photography. This park runs for about 250
yards west along the river and is about 70 yards wide. As
mentioned above, all birds were flying from the Yard with weights
by 17 February 1977. The first experiments were from across the
river in Anacostia Park (launch site one in Figure 19). At this site
the birds were "single-tossed- (launched one at a time) in an
attempt to collect individual data on vanishing bearings. However,
with few exceptions, the birds circled in the area for two to five
minutes and departed in various directions. The launch site was
moved across the river (launch site two in Figure 19) and then to
just east of the 11th Street Bridge (launch site three in Figure 19)
with only slight improvement in results. At this time it was decided
that the reason the birds were circling for several minutes was
because they were waiting for other birds with which they could fly
home. Therefore, on 9 March double-toss experiments (launched in
pairs) began at launch site three with a significant improvement in
results. Most pairs of birds took up an immediate heading in the
direction of home. This experiment was repeated on 10, 11, and 15
March with comparable results. Figure 30 shows a single-toss
experiment from launch site two.
FIGURE 30. Single-Toss Experiment
2,E�C�
1
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top �cret RUFF
�,) On 17 March, a detailed series of experiments began from
the 11th Street Bridge (launch site four). during which 84 paired
flights were launched. Visual landmarks were located with respect to
the museum park so that miss distance data could be recorded for
each pair of birds. The actual launch site was adjusted along the
bridge in order to maximize the likelihood of overflying the park.
Of the 84 flights from this site, 21 were recorded as hits (a miss of
less than 35 yards from the center of the park). Figure 31 is a
histogram of this data in terms of miss distance from the center of
the park. Since the bridge site is about 400 yards from the park, the
angular miss can be determined as the arc tangent of the miss
distance divided by 400. Twice the angular miss (measured from the
target center) can be considered as the total subtended angle of a
hypothetical target. In this way one can sum the data in Figure 32
to compute the cumulative distribution as a function of the
subtended angle of a particular target as measured from a chosen
launch site. Figure 32 shows the chance of overflight by one pair of
birds as a function of the subtended angle of the target. For
example, if the target in question subtends an angle of 25 degrees,
there is about a 50-percent chance of overflight. However, this
applies only when the launch site has been adjusted to locate the
most probable direction home through the center of the target.
751 For most of the launch sites in Figure 4, the Sudomekh
Yard subtends an angle of 10 to 70 degrees. This would indicate a
TCS 35974-77
35 55 85 135
Miss Distance (Yards)
185
FIGURE 31. Histogram of 84 Paired Flights
;85/
To ecret
235
Miss
34
SEERET
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
0.6
0.5
0
0.4
_o
0.3
0.1
I I I I I I I
00
10 20 30 40 50 60 70
Subtended Angle of Target (Deg)
FIGURE 32. Chance of Overflight vs. Subtended Target Angle
S T
25- to 60-percent chance of overflight once a preferred site (or
flight line) has been determined. If there is good visibility, the
direction of the preferred flight line can be established without risk
of flying the avian camera. Otherwise, as was the case with
Andrews AFB, the avian and satellite photography must , be, com-
pared to determine the new trial launch site.
feFej The 84 paired flights were conducted with eight birds, four
from Kit IF and four from Kit 2F (the young birds). During the
last ten flights, four pairs were recorded as hitting the target. Figure
33 shows a bird with camera returning from a flight over the Navy
Yard.
NPIC analyzed 36 rolls of avian film. Of these, six rolls
contained 23 frames of the museum park.
Relocation Experiments
) One of the first problems on this project was to find
competent loft managers who could provide and receive birds for
relocation experiments. It was felt important that these lofts should
be as widely separated as possible and that one should be in Alaska
which has the same latitude and magnetic dip angle as Leningrad.
Three highly competent people were found in Oregon;
Anchorage, Alaska; and Virginia. A fourth loft was
constructed in Missouri, on the farm of
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE 33. Bird With Camera Returning From the Navy Yard
1
ET
This fourth loft provided relocation
data by an -informed novice- as well as a ready supply of birds for
flight testing the avian camera prior to shipment.
-*$T Three kits of birds (1L, 2L, 1W) were obtained from the
area. One of these kits, 1W, had been relocated several
times_hefnre onihe previous project. A fourth kit, 1D, was obtained
from as young birds which had never flown. All other
birds were o ained from Oregon, Alaska, and Missouri, and
distributed to the four selected lofts. Table 9 shows the disposition
of the 118 birds used in this experiment, including the date each kit
was received and the date of the first release at the new home loft.
Table 10 shows the results for each of the 118 birds, including the
number of days each bird was held captive before the first release.
As expected, the birds which had never flown before, kit 1D, had
the highest percentage of survival (55%). This, however, is low for
young birds and most probably due to the fact that they were too
old and -wing strong- causing them to fly off before learning the
surrounding area. This is essentially the same percentage, and
probable cause, described previously for kit 2F. In Table 10, loss
data is shown by number of release at the loft (1R, 2R, etc.) and
also with respect to the number of training flights from a remote
launch site (1F, 2F, etc.). All birds that survived were trained from
sites at least five miles from the loft.
TCS 35974-77
Top ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
cfergVFF
TABLE 9
DISPOSITION OF 118 BIRDS
NUMBER
NUMBER DATES DATES DAYS
KIT BIRDS ORIGIN DESTINATION RECEIVED RELEASED CAPTIVE
1W* 6 California Oregon 10/25/76 12/26/76 62
IL 13 California Oregon 10/25/76 12/02/76 38
2L 12 California Oregon 10/25/76 12/05/76 40
1D** 11 California Oregon 11/30/76 12/26/76 26
IB 12 Oregon Missouri 11/05/76 11/27/76 22
2B 12 Oregon Alaska 11/06/76 12/08/76 32
3B 6 Oregon Virginia 11/05/76 12/23/76 48
IS 12 Alaska Oregon 11/09/76 12/26/76 47
2S 12 Alaska Missouri 11/09/76 11/28/76 17
1H 12 Missouri Alaska 10/30/76 12/07/76 37
2H 10 Missouri Virginia 10/29/76 12/17/76 48
*Had been relocated several times previously.
**Young birds which had never flown.
TABLE 10
RELOCATION EXPERIMENTS WITH 118 BIRDS
BIRD NUMBERS
KIT 1 2 3 4 5 6 7 8 9 10 11 12 13
1W IR IR * IR *
1L D IR * IF D IF IF 3R IR IF IR IR
2L IR IR H * D IR IR * * *
ID IR IR * * * * IF IR * D
IB 3F IR 2R * 2R 2R IR 2F 2R * * *
2B 4R 4R * 4R 4R 4R 4R 4R 4F D 4R
3B IR 17F 2R IR IF
IS IR I H S 2R * I * * 2R
2S 2R * 2F 3R 3R 2R * 3R 3R 3R 2R
IH 4F 4F 4F 5R 5R 5R 9R 5R 5R 4F
2H IR D * E I D * D 15F
D=died, I=injured, S=sick, H=hawk, E=escaped.
IR, 2R, etc.=lost on 1st, 2nd, etc., release at loft.
IF, 2F, etc.=lost on 1st, 2nd, etc., flight from remote site.
*=survived and homed in (31% of 118 birds).
NOTE: Kits 1W and ID were not 1st relocation birds.
A Of the 101 birds (discounting kits 1W and 1D) being
relocated for the first time, Table 11 shows survival statistics relating
to the origin of kits, and Table 12 shows statistics relating to their
destination or new home loft. The low origin statistics for Missouri
and Oregon (23%) are primarily due to the low destination statistics
for Alaska (only 17%). In order to investigate this anomaly, which
could impact severely on the proposed Leningrad targets, an
additional kit (3B) of 25 birds was obtained from Oregon and sent
to Anchorage on 1 January 1977. One of these birds died. Of the
remaining 24 birds, 12 were flown many times from 30 miles, and
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top S et RUFF
TABLE 11
ORIGIN STATISTICS ON 101 BIRDS
ORIGIN
KIT
FRACTION
ORIGIN
FRACTION
ORIGIN
% SURVIVAL
California
IL
2/12
8/25
32
2L
6/12
Oregon
1B
4/12
7/30
23
2B
2/12
3B
1/6
Missouri
1H
2/12
5/22
23
2H
3/10
Alaska
1S
5/12
8/24
33
2S
3/12
TABLE 12
DESTINATION STATISTICS ON 101 BIRDS
1
DESTINATION
KIT
FRACTION
DESTINATION
FRACTION
DESTINATION
% SURVIVAL
Alaska
2B
2/12
4/24
17
1H
2/12
Oregon
IL
2/13
13/37
35
2L
6/12
1S
5/12
Missouri
2S
3/12
7/24
29
1B
4/12
Virginia
2H
3/10
4/16
25
3B
1/6
two of these have been to 120 miles. These 12 surviving birds have
been worked steadily (at least once a week) from 30 miles since
their release in February. At this writing, all 12 are still flying at
the Anchorage loft and have been worked harder than any other
kit.
...(S-} Table 13 shows overall relocation statistics for the 101 birds
with a 28-percent survival from losses of all kinds. Note there were
not always losses on the very first release. However, of the 44 birds
lost during release at the loft, 31 were lost the first time losses
occurred, and the remaining 13 were lost the second time that losses
occurred for each kit. This implies that losses do not always occur at
the first opportunity, but they do appear to occur in large groups. It
was noticed that the -first loss- occurrence did tend to happen on
bright sunny days with low wind and few clouds. The same
statements can be made with regard to the 15 losses during flight
training from remote launch sites; 13 birds left the first time losses
occurred, and the remaining two left the second time losses
occurred. It is interesting to note that of the 59 birds lost during
release and flight training, only 25 percent were lost in flight
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
training whereas over half of the losses happened the first time
losses occurred.
A small project loft see Figure 34) was constructed adjacent
to the loft at Virginia, which was used to
train kit 2F to fly, an a so for the relocation of kits 2H and 3B.
Table 12 shows that four of the 16 birds survived relocation.
However, Table 10 shows that two of these birds (2H10 and 3B2)
had made 15 to 17 flights (4 or 5 with weights) before they were
lost. Table 14 shows the flight performance for these two relocated
kits in terms of the number of times they carried weights of the
same shape and weight as the avian camera. Bird 2H8 was
eventually lost in April 1977, after numerous flights from the Navy
Yard, and was not counted as a relocation loss. Two of these birds
were moved to the widowhood experiment to be discussed later.
Additional Relocation Experiments
48 Including kits 1W and 1D, there were 33 birds surviving
the relocation experiments in Oregon, Alaska, and Missouri. Most
FIGURE 13
FIRST RELOCATION STATISTICS
NUMBBR
OF BIRDS
LOST IN RELEASE
AT LOFT
LOST IN FLIGHT
TRAINING
KIT
Start
Finish
Died,
etc.
IR
2R
3R
4R
5R
>5R
1F
2F
3F
4F
5F
>5F
IL
13
2
2
4
0
1
4
2L
12
6
2
4
1B
12
4
2
4
0
I
2B
12
2
1
0
0
0
8
0
0
0
3B
6
I
2
1
1
0
0
0
0
1
IS
12
5
4
1
2
2S
12
3
0
4
4
0
I
IH
12,
2
0
0
0
0
5
1
0
0
0
2H
10
3
5
1
0
0
0
0
0
1
Subtotal
14
11
5
8
5
1
5
2
1
5
0
2
Total
101
28
14
44
15
TABLE 14
FLIGHT PERFORMANCE FOR RELOCATION KITS 211 AND 3B
BIRD NUMBER
TOTAL FLIGHTS WEIGHTS NAVY YARD WEIGHTS
2H2**
7
3
2H4
21
15
2H8
19
13
2H10
3
o
3B2*
4
0
3B6**
7
4
*Lost after 15 to 17 flights.
**Moved to widowhood experiment 3/3/77. 2H8 was eventually lost 4/11/77.
1
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
FIGURE 34. Exterior View of the Attic Loft
UNCLASSIFIED
loft managers typically train their birds to come into the loft
quickly in response to audio cues, such as whistles, rattling feed
cans, or door chimes. During these experiments, each loft manager
was provided a magnetic tape cassette with a particular music
selection that would not normally be associated with an audio cue.
All birds learned quickly to respond to the musical cue. In addition,
all birds learned to enter the loft through an air-conditioning shell
which had a 41/2-inch hole in its top face. Exit from the loft, for
exercise, was provided by emplacement of a ramp leading up to the
drop hole.
The surviving 33 birds were shipped to Virginia by air
freight. Four were used in the widowhood experiment (to be
explained later) and 29 were placed in an attic loft. Figure 34 is an
exterior view of the attic loft. The top most air-conditioning unit is
a shell complete with drop hole and entrance way from the attic as
shown in Figure 35. In Figure 34, the view is blocked to the south
and west by the apex of the roof. To the north and east the view is
also blocked by tall trees. The birds were held captive about 40
days during which time a round of youngsters were raised. Also,
reinforcement to the music and drop hole was provided by using a
simulated air-conditioning shell located on the loft floor; this
training is shown in Figure 36. The 29 birds were organized into
two groups. For the first release, the first group of 12 birds was
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
FIGURE 35. Interior View of the Attic Loft
UNCLASSIFIED
FIGURE 36. Reinforcement Training to Drop Hole
UNCLASSIFIED
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
allowed to leave the air-conditioner shell by use of a ramp similar to
that shown in Figure 37. All 12 birds were lost on the first release.
It was speculated that the birds could not see enough of the
surrounding area through the louvers in the air-conditioning unit to
prevent being lost. Therefore, the second group of 17 birds were
hobbled by placing rubber bands around the last four or five
primary flight feathers. This causes a gross aerodynamic imbalance
and greatly suppresses the ability and desire to fly. On the first day
of release, 10 of the 17 hobbled birds walked out onto the roof and
spent the day. Six of these birds did not reenter the loft that
evening, but all were in by the next day. After four days, all 17
birds had spent between two to four days on the roof and the wing
hobbles were removed. On the first release without hobbles, 9 of the
17 birds were lost. The tenth was lost on the third release, and the
eleventh was lost on the fifth release. On the seventh release, a cat
was observed on the roof chasing the remaining six birds and all
spent the night away from the loft; one of these did not return. By
the end of the eighth release, there were five birds left, and flight
training from remote sites began. By the end of the eighth flight, on
14 May 1976, all five birds were flying several miles back to the
loft. A survival rate of five in 17 (29%) is typical of the previous
experiments. At this time, however, unseasonally hot weather moved
into the Washington area with temperatures in the mid-90's. The
temperature inside the loft climbed to well over 100 degrees despite
the use of overhead insulation and the installation of an attic fan in
FIGURE 37. Bird Leaving Air-Conditioning Shell
CONUENTIAL
63
TCS 35974-77 Top
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
the loft area. On the ninth flight, three of the five birds did not
return. The two that did were both hens that were sitting on eggs.
At this point, the experiment was terminated because of the
excessive temperature in the loft. It is apparent that there are a
multiplicity of factors affecting relocation. Some of these are
discussed in more detail in Appendix C.
Widowhood Experiments
je) The widowhood system is a complicated technique used by
the more experienced pigeon handlers for the purpose of highly
motivating cocks to return to the loft quickly and, thereby, win
pigeon races. Some of the particulars are described in Reference 4
and in Appendix C. Briefly, cocks are taught that they may be with
their hens only when they are taken to a remote site and released to
come- home. When a cock is working well on this system, he comes
straight home and immediately enters the loft to be with his hen.
Of the 33 birds transported to the Washington, D.C. area for
additional experiments, four were placed on widowhood to study the
effects on relocation. These birds were 2L5, 2L11, 1B11, and 1S11.
Bird 2L5 was lost at the loft for unknown reasons. Bird 2L11 was
relocated successfully and flown three times with weights, the last
time from the Washington Navy Yard. Bird 1B11 was lost during a
severe thunderstorm, and bird 1S11 was relocated but finally lost
with a weight while flying back from the Navy Yard. Two of the
old birds, 1F5 and 1F14, were placed on this system and did carry
weights from the Navy Yard. The last, a young bird (2F10), also
carried weights from the Navy Yard. Figure 38 shows the widow-
hood loft with air-conditioner shell for the entrance of the four
relocated birds. Two birds, 2112 and 3B6, which had survived the
first relocation experiments, were tried on this system but failed to
perform properly. The general consensus at this time is that the
widowhood system is too complicated to be used in the field. At
some future time it might be reconsidered as an advanced technique
to be used by highly trained loft managers.
Remarks on Behavior
>SI It is apparent that the birds are capable of getting the
camera over targets such as Andrews AFB and the small museum
park in the Washington Navy Yard. By applying the results of
Figure 33 to the problem of Sudomekh in Figure 4, it would appear
that there is a 25 to 50 percent chance of overflight once a
preferred flight line- has been established. The fact that homing
pigeons do not always have a vanishing bearing precisely in the
direction of home is referred to as "launch site bias- and is
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
FIGURE 38. Widowhood Loft With Air-Conditioning Shell
UNCLASSIFIED
discussed in Reference 5. The reason for this bias is -not clearly
understood and must be measured for each general locality. Refer-
ence 3 also discusses 'visual cues wherein homing pigeons will use
tall objects near their home loft to which they will visually home as
a -terminal guidance- object.
The relocation of homing pigeons has been a subject of
study for some time. An excellent book (Reference 6) by Dr. W. E.
Barker lists several essential points which are discussed in Appendix
C. One of his points states that if relocated birds are placed under
stress, they may well decide to leave. Certainly, this tendency was
experienced ,on�this program. On the other hand, the U.S. Army
relocated many pigeons during World War II. The exploits of
several decorated pigeons are described in the first chapter of
Reference 7. Here, pigeons carried messages through storms and
enemy gun fire, many returning with severe wounds. Certainly,
these could be considered as conditions of severe stress. However,
the precise techniques and methods used for relocation, and the
survival statistics, have not been thoroughly researched and defi-
nitely should be, if further work is to be done on this project.
Relocation experiments were conducted on a previous
project during the spring and summer of 1976. Here, the entire
pigeon loft (a packing crate) was transported to the new location
site. In this way, the birds already knew what the outside of their
TCS 35974-77
Top
)os5c,
cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
new home looked like (a primary point in Barker's book). During
the -attic loft- experiment, the second group of birds learned the
view from outside the loft by the use of -wing hobbles,- which
appeared to work well. A still better way, if operationally feasible,
was used by a loft manager in Japan, wherein he placed his
imported birds in a small cage on the back of his Honda and drove
them around the neighborhood.
f,S-r In conclusion, the reliability of relocated birds is still a
serious question, regardless of the method used for resettling.
Without further research on reliability, it is recommended that
young birds be taken to the target loft so that this will become their
primary home. They can then be relocated to the Washington area
for training and selection. Those of acceptable performance can then
be returned to the target loft with a high degree of confidence that
they will perform with essentially the same statistics as that
described in Figures 31 and 32. During the experiments conducted
in the Washington, D.C. area, there were 51 flights with cameras
and 341 with simulated weights, or 392 flights with cameras or
weights. These data indicate that there is about one chance in 30 of
losing a camera on each flight. However, in about half of these
losses, the bird and harness returned to the loft. A better method of
camera attachment would reduce losses to about one in 60.
TCS 35974-77
6K19 T ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top Se RUFF
PART 6
AN EXAMPLE MISSION SCENARIO
(0) Several assumptions are necessary. For example, assume we
wish to complete a mission in the Leningrad area before the end of
the 1978 calendar year. Figures 39 and 40 show that only from the
Tth
a)
50
40
10
June
May
Example
Date 1 March
Elevation 20�
Hours Above
20� 3
April
(See Dashed Line)
March
�I
I
�
February
i
Ja"-------....\\nulary
I
I
i
I
5 10 15
Hours
FIGURE 39. Hours the Sun is Above a Given Elevation
for January -June in Leningrad
UNCLASSIFIED
67
20
TCS 35974-77 Top
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Tw1ecret RUFF
a)
C/)
5
4
30
20
10
0
Au.ust
11111
Example
Date 1 October
Elevation 2()�
Hours Above
20(See
Dashed Line2
)
eptember
October
11,1m
k
.... November
NDecember
0
5
10
Hours
15
20
FIGURE 40. Hours the Sun is Above a Given Elevation
for July - December in Leningrad
UNCLASSIFIED
latter part of February to the early part of October does the sun
rise above 20 degrees in elevation for any appreciable period of
time during the day. Results from the balloon tests in Part 3
indicate a sun elevation of at least 20 degrees is necessary for
adequate contrast and exposure. It is felt that operational flights
against the targets in Leningrad should begin no later than early
August 1978 so that adequate coverage can be obtained by early
October. Several plans are possible which trade risk against oper-
ational complexity. Two plans are discussed below which assume a
start date of 1 October 1977.
TCS 35974-77
68
ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
Plan I (preferred)
(IR) A local loft must be purchased or constructed by mid-
October 1977. During this time, 100 pairs of exceptional breeders
will be purchased at an average cost of $150 a pair. These will be
placed in the local breeding loft by the third week in October so
that all will be on eggs by the end of the first week in November.
These eggs (about 150 to 170) will hatch by the end of November.
By the end of December, about 30 days of age, all young will be
out of the nest eating and drinking on their own. During these three
months, a loft manager must be trained in the care, handling and
flight training of young birds, and a loft must be constructed in the
About half of the young
birds (70) will be handpicked for transport to
the first week in January 1978; the subject of transport will be
discussed later. During January these birds will be trained into the
as their primary home. If possible, the birds should be
placed just outside the roof access window, shown in Figure 6, in a
wire basket for several hours a day just before their first release
(about one week after arrival). Certainly, the basket should not be
placed where it can be seen by any possible observation post. If the
basket cannot be placed on the roof, the birds should be allowed to
look out as many windows and dormers as possible prior to their
first release. The birds can be released in small groups which are
kept hungry so that they will come in quickly when the dinner
music is played. By the end of January, or the first week in
February, all birds should be flying from the loft and ranging a
mile or so during exercise flights. There will probably be about 60
During the remainder of February, the birds
in small groups and released in parks or
will be said later on clandestine release
not mandator, 11-Fh rsf vebruary,
will consider as their
birds left at this time.
should be taken out
wooden areas (more
techniques) but this is
about 50 young birds
primary home.
..(1 There is certainly a temptation to select some of these 50
birds for the purpose of testing their performance with weights in
the Leningrad area during the month of March and trying for a
mission in April or May. However, this may involve undue risk and
would require sending a new loft manager to the field who had
been trained in mission-oriented techiques. A safer approach would
be to bring the 50 birds back to a loft in the Washington, D.C. area
By mid-April, these 50 birds will be ready for their first
release from the local loft. Relocation techniques not available in the
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top icret RUFF
field can be used to increase the relocation rate of survival. It is
estimated that about 20 to 25 birds will survive this relocation. By
the first of May, flight training to harness and weights can begin,
and the birds can be moved out toward local example targets (e.g.,
the Washington Navy Yard). Through the end of June, field
personnel will be trained in all aspects of clandestine release, camera
maintenance and repair, and techniques for determining the -pre-
ferred flight line.- By the end of June 1978, 10-15 birds will be
selected for transport back to the Leningrad loft.
...4-T57 During July, these 10-15 birds, which consider Leningrad
their primary home, can be used to determine the preferred flight
line and to select suitable launch points for overflight of the selected
targets. By the first of August 1978, mission operations can begin
with some confidence that the probabilities of overflight are
essentially those of Figure 32. The disadvantages of this plan are:
1. There are only about two months before operations must be
terminated due to low sun elevation.
2. The birds must be transported three times.
3. There is a great deal of -activity- in the field
4. The time schedule is very tight with some risk of being late.
Plan 2
..4P5)Thuring the first half of October an operational type loft is
constructed in the Washington, D.C. area capable of housing 200
pigeons. These birds will be two to three years old and all should be
of proven worth from their racing records. The cost will be in the
neighborhood of $100 each. These birds will be relocated to the
local loft for a first release by the early part of December 1977.
Again, techniques not suitable in the field can be used to increase
survival. All birds should be on eggs by the first release, and wing
hobbles should be used for one week after the first release. It is
estimated that there will be 70 to 100 surviving to the end of
December. Starting in January 1978, these birds will be trained to
harness and weights and worked out to example targets. By the end
of February, loft managers will be fully trained in all aspects of the
mission, and the final selection of 35 to 50 birds will be made.
These birds will be shipped to Leningrad in early March and the
first release, with wing hobbles, will occur during the end of April.
Again, all birds should be on eggs at the time of release. At least 10
to 15 birds should survive to the end of May. The performance and,
hopefully, the reliability of these birds have already been established
over local example targets. By early to mid-June, launch sites should
TCS 35974-77
T ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top ecret RUFF
be selected and mission operations can begin. The advantages of this
plan are:
1. Operations start in mid-June giving about four months to
collect photographic coverage.
2. Only 35 to 50 birds are taken to the field once, and all these
birds have been tested over local targets.
3. There is no need to rush into the field, as with plan 1, until
all personnel (and birds) are fully trained.
The disadvantages of this plan are:
1. The care and handling of 200 pigeons through the first
relocation is no easy task, not to mention the harness and
weight training of 70 to 100 birds.
2. The estimate of 35 to 50 birds of acceptable performance for
shipment to Leningrad is speculation. There is insufficient
data from this research for a reliable prediction.
3. Assuming there are a sufficient number of birds which
perform well over local targets, the probability of their
holding up under additional stress at the operational target is
a matter of conjecture at this time. Without further research,
it must be assumed that the risk of loosing camera and bird
is higher with this plan than with plan 1.
.61 Plan 1 is more complex but involves less risk in predicting
performance. With additional research on the performance of birds
at a secondary home, plan 2 may well be preferred; it should
certainly be the goal of further research.
Transportation
,(3") During this project birds were shipped by airfreight from
Anchorage, Alaska, to Dulles Airport on several occasions. During
one shipment from Oregon, birds were lost for three or four days,
finally arriving in fair condition; the primary danger is thirst and
heat, not hunger. It is not uncommon for local loft 'keepers to ship
birds to Japan or receive them from France and Belgium.
fe The issue here is that of clandestine transport. Some
research has been done in this area (Reference 8) but with species
other than pigeons. However, it is felt that a pigeon in good
condition is among the hardiest of bird species and should survive
transport as well as any. In Reference 8, birds were transported in
-hole luggage- and -carry-on luggage- for periods up to 72 hours.
Some holes and air passages (quarter-inch holes in the bottom) were
provided for ventilation. The birds were wrapped in cloth, or panty
TCS 35974-77
71
Top ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
lop cret RUFF
72
TCS 35974-77 Top S
hose, to restrict movem t, and eye coverings were used which
greatly inhibited auditory noise and caused a -catatonic- like state
of inactivity. A layer of San-O-Sil (trademark) impregnated with
water-soluable neomiocin was placed in the bottom of the luggage
in order to prevent a toxic contamination of the air by ammonia in
the birds excrement.
Clandestine Release
Most are familiar with the magic tricks that seem to
produce large numbers of birds from nowhere. In fact, birds
transport well in the outside or inside pockets of large overcoats.
Consider a large inside pocket near the bottom of an overcoat. If
one were kneeling, say to feed the park pigeons, it would be a
simple matter to release several birds from underneath the coat.
One particular method was studied which would be
convenient for the release of a number of birds over a short period
of time. An access hole was provided in the floor of an automobile
through which one or two pigeons could be placed onto the road or
parking lot. The pigeons walked from underneath the car and
immediately took flight upon reaching daylight. Figure 41 shows a
pigeon with harness and weight that has just taken flight after being
released by this technique. If the automobile were in a parallel
parking area with adjacent parked cars, it is conceivable that several
birds could be released by this method, even under close surveil-
FIGURE 41. Pigeon In Flight After Clandestine Release
2it,tSE T
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
lance. Various methods ranging from shopping bags to briefcases
with -trick bottoms- could be tried in the local area during training
exercises.
The Loft
...tej The loft in the target area must be completed prior to
arrival of the first birds. This includes perches (a collapsible egg
crate frame of 1/4 in. plywood), nest boxes (if they are old birds),
and an adequate supply of feed and medicine (see appendix C for
the prevention and cure of disease). A grown pigeon will eat about
an ounce of feed per day, and a few pounds of grit will last a
month for 50 birds. The loft must be well ventilated without
noticeable drafts. The colder the temperature, the better the birds
thrive (the Alaskan loft has experienced temperatures of 40 degrees
below zero). However, high temperatures (above 90 degrees) will
precipitate disease and severely degrade performance. If loft tem-
peratures above 90 degrees are to be expected (at the target area or
locally), some combination of insulation and air-conditioning must be
provided without the birds being in a direct draft. This issue cannot
be taken lightly, particularly with young birds.
1
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top ret RUFF
PART 7
SUMMARY AND CONCLUSIONS
.4S-) It is felt that the quality of photographs collected over the
Washington Navy Yard is adequate for satisfying a high-resolution
requirement. Furthermore, the chance of overflight, shown in
Figure 33, certainly demonstrates the feasibility of getting the
camera over the target. An unexpected result of this ten-month
effort was a lack of data sufficient to predict the performance of
relocated birds. This is due primarily to the initial allocation of
priorities dictated by the constraints of manpower and funds.
Photographic Quality
(S/TK) Of 36 camera flights over the Navy Yard, six collected
23 images of some portion of the target. Mensuration on 53 samples
varying from 3 inches to 37.5 feet showed an average error of less
than 1.6 inches. Perhaps of greater interest is the mensuration of 22
small objects (less than 2 feet) which showed an average error of
about 3/4 of an inch with 90 percent of all errors less than 1.8
inches. In comparison with KH-8 photography of the same target,
the avian system was rated as having a higher image interpretability
as well as the ability to see smaller objects. The National Imagery
Interpretability Rating Scale (NIIRS) ratings were 7.8 for the avian
system and Furthermore, it is believed that the avian
system, due to its extremly low altitude, has a high potential for
using color photography (SCAT) for materials identification (see
Table A4 in Appendix A).
Camera Research
8'rIt is felt that the camera can be brought from
research to operational status within the time frame of either plan
discussed in Part 6. However, it is suggested that the following be
considered carefully:
1. The field replaceable shutter assembly has greatly contrib-
uted to system reliability. However, a method for measuring
shutter speed in the field should be developed as a check on
system performance. Low shutter speeds were the primary
cause of image blurring on this project.
2. A method should be developed for identifying the type,
location, and degree of light leaks prior to each mission. This
TCS 35974-77
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
To1et RUFF
3.
problem is not difficult to solve and would provide
important check on system performance.
An effort should be made to reduce the weight
system, particularly the method of harness
ment which is subject to periodic failure. Research
area would improve bird performance.
a most
of the
attach-
in this
4. Some barrel distortions in the lens were observed
at the edge of the field of view. Mensuration accuracy could
be improved if the lens were calibrated or redesigned.
5. Appendix B contains low temperature data and suggestions
for providing a low temperature system.
.481 Some preliminary research was conducted to increase the
lateral coverage by use of a miniature panoramic camera which has
the same weight as the current system. The lateral
coverage is adjustable and would increase the oblique photography
to the right and left of the flight path. Figure 42 is a picture of this
FIGURE 42. Miniature Pan Camera
TCS 35974-77
76
Top ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
FIGURE 43. Pan Picture With 180� Field of View
CBINTIAL
miniature pan camera. Figure 43 is a picture taken by this first
research unit with a 180 degree field of view. The goal would be to
achieve the same photographic quality as the current avian system.
However, no further research is planned at the present time. The
point here is that increased coverage would provide increased
intelligence and, therefore, fewer flights and less risk.
Bird Performance
...481 The ultimate goal should be to provide a hard core
operational kit of birds which have been trained over local targets
and selected for proven performance. These birds would be held at
a local operations-type loft for rapid deployment to any target area
with confidence in the expected performance and reliability. It is
felt that the key ingredient in achieving this goal is the local loft
and the dual involvement of research and operational field person-
nel. Aside from providing a training ground for personnel and a
proving ground for new ideas, such a loft might act as a vehicle for
TCS 35974-77
77
Top ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
cret RUFF
fusing together the practical problems in the field with the science
of re'search.
Concluding Remarks
r Whether or not sufficient data has been presented to
warrant the initiation of an operational plan is dependent on the
importance of the intelligence need in relation to the risk and
logistic complexity in the field. It is hoped that this report contains
sufficient data and candor to allow those concerned with the
intelligence requirement and field operations to form a proper
judgment. It is suggested that an active interchange between
research and the concerned parties may provide a plan with
acceptable risk and logistic complexity, and clearly identify the type
and degree of research required for adequate support.
1
TCS 35974-77
78
Secret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
1
CI
APPENDIX A
EVALUATION OF PHOTOGRAPHIC COVERAGE
(Prepared by NPIC)
Section 1 Summary
1.1 Introduction
During the past year (6/76-7/77) the National Photographic
Interpretation Center has provided analytical and production support
for a research study (Tacana) being carried on by the Office of
Research and Development/Operations Technology Division to de-
termine the feasibility of using an avian (pigeon) platform for
purposes of collecting overhead reconnaissance photography. This
support has included:
� The evaluation of the system imagery for exploitation
purposes by the Imagery Exploitation Group (IEG) and the
Technical Services Group (TSG).
� Assistance in the camera modification development, film
selection, and system quality evaluation by the Technical
Services Group.
� Imagery production and film processing by the Production
Services Group (PSG).
(U) Following is a summary of the results of the various studies
and observations which will aid in describing the quality of the
exploitation product from this system.
1.2 Interpretability
(S/TK) The Tacana system has the capability to acquire
imagery of an exploitation quality equal to, or better than
existing systems, including the KH-8.
1.3 Mensuration
.4.$74-11C1 The Tacana system can image measurable objects of
small dimensions. Twenty-two measurements made of object di-
mensions of 3 inches to 2 feet indicate that 90% of the errors (from
ground truth) are less than or equal to 0.15 feet (4.5cm). Fifty
percent of the errors are less than or equal to 0.06 feet (1.82cm).
79
TCS 35974-77 Top S
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
1.4 Digital Image Manipulation
_(..S< Interactive Digital Image Manipulation techniques
clearly reduce the imaging effect of directional smear some-
times produced by the Tacana system. The beneficial effects of
other IDIMS techniques (to enhance the detail of shadows and
highlights, reduce grain noise and clarify fine detail) were less
obvious. These techniques will usually benefit any general film
imaging system (including Tacana) that has acquired threshold level
detail imagery (where density and resolution information is almost
present).
1.5 Image Quality Analysis
The Tacana system has the potential to acquire imagery
of less than 1 inch (2.54cm) GRD under optimized conditions of
lighting, platform stability, camera reliability, and flying height.
Most of the time the system will perform at a lower quality level
due to the unpredictable nature of the platform and variable flying
heights. A camera shutter speed of at least 1/2000 second is
required to produce a reasonable percentage of sharp (non-smeared)
images from a given flight.
1.6 Film Selection
% The Tacana is a fair weather (with sun) system and is
adaptable to both black-and-white and color films (EK 3410, FE
6526 and Aero Color Negative 2445). Because the camera is a fixed
exposure unit, clear sun (no clouds or heavy haze) acquisitions of
less than 20 degrees solar elevation will require a faster film with a
resultant loss in overall image quality.
1.7 Other Summary Comments
1.7.1 j�S�Y An avian (pigeon) platform can be used to
acquire imagery of a specified target area. Of 36 test flights
evaluated over the Washington Navy Yard of a designated target
(Willard Park, the Navy Museum display area of about 64,000
square feet), 6 flights acquired 23 images of some portion of the
park.
1.7.2 (U) Throughout this study, image quality was com-
promised due to poor camera reliability (i.e., power failures,
shutter breakdowns, light leaks producing fogged film, optical
system misalignments and/or internal flare). While these -kinds" of
malfunctions and effects are typical of problems associated with
research and prototype fabrication, a clearer view of system
performance could be obtained if they were corrected in future
development efforts.
1
TCS 35974-77
80
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
Top ret RUFF
1.7.3 (U) The pre nt lens produces imagery
containing a high degree of barrel distortion. The mensuration
accuracy could probably be improved if each camera was calibrated
and future lens systems (i.e., design) had reduced distortion effects.
1.7.4 >rrGreatest detail imagery will be obtained with this
system on black-and-white film.
1.7.5 �(-St Using color films this system has a high potential
for materials identification (using SCAT) and certain camou-
flage, concealment and detection applications. Further research
should be done to determine the effectiveness of this system in these
areas. The Tacana's large scale imagery and its relative freedom
from high altitude atmospheric effects during acquisition are benefi-
cial for this type of analytical analysis.
Section 2 Determination of Exploitation Suitability
2.1 Introduction
.4..S74511Z) Because the Tacana study was a research effort, it is
felt a measure of performance from a user point of view would be
meaningful to aptly describe the quality of the system product.
Following are NPIC evaluations that describe the exploitation
(interpretation, mensuration, and image manipulation) potential of
the system. The nature of this study involved parallel development
of the bird, camera, targeting philosophy, and films and processes;
therefore, it is impossible to predict a level of system performance
to be expected operationally. However, efforts were directed towards
describing the exploitation suitability of the system in terms of its
potential to perform when all elements appear to be functioning at
their optimum, i.e., no camera malfunctions, tired and untrained
birds, or wrong film/poor process combinations. Certain KH-8 data
(and photography) is included in the analysis for informational
purposes. For editorial convenience, a study describing the capability
of the Interactive Digital Manipulation System to improve the
Tacana exploitation product is placed in this section.
2.2 Interpretability
2.2.1 Figure Al shows the percentage occurrence of
NPIC/IEG photointerpreter NIIRS ratings* of 85 images acquired
LSYNHRS (National Imagery Interpretability Rating Scale) is a uniformly
understood and systematically applied judgment by photointerpreters of the
interpretability of acquired imagery, regardless of collection source. It is a
graduated scale divided into 10 numbered rating categories, with 0 representing
unusable imagery and nine representing imagery with the best interpretability. It
is enough to know here that NIIRS is an accepted Intelligence Community
measure of interpretation quality and is used to aid collection and mission
planners, engineers, photoscientists and other PIs. For further information see
NIIRS Documentation, Vol. II, TCS-9842/74.
81
To cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top et RUFF
by the Tacana." system over the Washington Navy Yard 3/17/77.
This data represents an average rating by three PIs of a complete
flight operation. Operational PI NIIRS data from KH-8 Mission 4348
is also shown. The Tacana flight evaluated represents the best
performance of the total system (in terms of image quality) that was
obtained during the test phase of this study. This is not a measure
of predicted system performance but does describe the capability of
the system to produce high-quality imagery when all aspects (bird,
camera, weather, etc.) are performing reliably.
2.2.2 /) Figures A10-A17 represent Tacana vertical and ob-
lique images from the flight that were rated by the PIs. Other types
of photographs typically acquired by this system are illustrated by
Figures A18 thru A20 showing observed maximum flying height
(about 300 feet) and Figures A21 and A22 representing adjacent
frame quality differences caused by inconsistent bird motion (a
common occurrence with this system). Figures A23 thru A27 show
sequential (adjoining) frames from one flight covering the Navy
Yard, Willard Park Display Target Area. Of interest here is to note
the target mapping effect that can be accomplished with this kind
of a system. It is useful to compare these pictures with the KH-8
coverage of the target area (Figure A30). Figures A23 thru A27 also
exhibit certain camera anomalies that are more completely described
in Section 3.2.7.
2.2.3 (U) There are no peculiar problems associated with
viewing Tacana imagery. Its small format image (12.5 x 12.5mm)
and large scale (nominal 2000:1) lend themselves well to producing
duplicate positives, handling, and viewing with existing (low magni-
fication) photointerpretation equipment.
2.2.4 .(�).....T-ICT Due to the small area coverage of each frame
(nominal 30 x 30 meters) and the inconsistent aiming of the
platform, some type of collateral coverage should be available to the
interpreter for determining his Tacana coverage in the general
target area. KH-8 coverage is suitable for this mapping function.
2.2.5 (U) To simplify the duplication of this type of imagery
(16mm strips, nominal 3 meter lengths per 200 exposures) a three
density level exposure of each image segment will satisfy most
frame-to-frame density differences caused by ilkimination and target
brightness variations. For example, one image segment of 10-15
images would be duplicated three times (once each at three
different density levels) side by side on a 20.3 x 25.4cm (8 x 10
inch) chip of duplication film. This technique allows the PI to easily
select the best frame for interpretation as well as giving him the
opportunity to exploit two different brightness areas (shadow and
highlight) of a given scene.
TCS 35974-77
82
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top S et RUFF
2.3 Mensuration Suitab. ty
2.3.1 Introduction
(,511`K) An exercise was performed to provide a view of the
mensuration of imagery produced by the Tacana system. Supportive
data and accuracy standards were provided from the KH-8 and
ground truth information.
2.3.2 Exercise Procedures
.(4) A target test area similar to an operational area was selected
at the Washington, D.C. Navy Yard museum (Willard Park, Figure
A30). Within this test area, a TALOS missile was singled out as
being representative of the type of target appropriate for study
(Figure A29). This missile has an assortment of different shapes and
dimensions that would be both quantitatively and qualitatively
important in an intelligence sense.
(U) The procedures for this exercise were to first acquire
distances from the target site for use in scaling. These distances
ranged from 15 to 40 feet (5 to 12 meters).
(U) Two system photographs were used for this study (Figures
A25 and A28). Other coverages of the target area were not included
because of severe blurring due to platform motion, lens distortion
effects and/or extreme obliquity.
a. Scale Determination
Good satellite coverage of the Washington, D.C. Navy
Yard is minimal. Two KH-8 frames were found that could be used
for determining a working scale. Monoscopic measurements were
made and compared with ground truth dimensions. From these
values, one or more were chosen as a scale reference in deriving
dimensions from system imagery. The scale dimension used had to
be as close as possible to the target, generally parallel to the target
and nearly in the same object plane. The last requirement could not
always be done, especially in trying to get a variety of dimensions
from the TALOS missile which not only rests upon a raised base,
but also has a nose-up attitude. Because of this target problem,
additional targets in the vicinity of the TALOS were chosen in
order to demonstrate the capability of the system to -see- small
dimensions.
b. Focal Length and Format Edges
Although the cameras have unique design features for their
size and function (moving film and an image motion compensator),
TCS 35974-77
83
Top ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
the imagery can be analyzed as frame photography since the
photograph is exposed simultaneously over the entire format. With
that in mind, the frame format edges and focal length were looked
at to find out if the imagery lended itself to analysis by either
graphical or analytical means.
2.3.3 Results
Preliminary results* showed that small dimensions could be
measured. This exercise was to test the Tacana system for mensura-
tion feasibility on -operational- imagery which included long
dimensions (2-38 feet) and small dimensions (0.2-2.0 feet). The
results also compare the image formats of the two types of cameras
and their respective focal lengths. This additional information is
important for future exploitation of operational imagery.
a. Format Edges
fe81 The format edges are important in that they are used to aid
in computing the focal length. There are two types of cameras: one
with a rectangular format measuring 12.5mm by 7.5mm, and the
other with a square format 12.5mm by 12.5mm. The rectangular
imagery acquired with the Model I (MCW-24) had fuzzy format
edges when viewed on a monoscopic comparator at 5X magnifica-
tion. However, the square format imagery (Model II
camera) had sharp, well-defined edges when viewed on
comparator.
the
b. Focal Length
lee The nominal focal length of the Tacana camera is reported
as 15mm. A test was done to check this value for the rectangular
format camera. Two focal lengths were computed: 14.05mm and
14.19mm.
c. Target Data
IrT-871eIZ) Mensuration data was gathered from two photographs
(Figures A25 and A28). The data points were measured on the 829
Mann Comparator. There was a total of 53 dimensions measured
ranging from The scale was determined from
KH-8 measurements and the scaled distances compared with ground
truth. The errors, or differences from ground truth, ranged from
This data is listed in Table Al. Histograms were
*The preliminary tests were done using imagery taken from a fixed or static
camera platform, i.e., a balloon. No written report was required; however, results
can be reviewed.
1
TCS 35974-77
84
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top - -cret RUFF
TABLE Al
MENSURATION DATA FROM NAVY YARD IMAGERY*
1
GROUND MEASURED
TRUTH DISTANCE
ERROR
GROUND MEASURED
TRUTH DISTANCE
ERROR
* All dimensions given in feet
constructed for three separate data sets: overall dimensions, dimen-
sions less than or equal to two feet, and dimensions less than or
equal to one foot. See Figures A2, A4 and A6, respectively.
Corresponding graphs showing cumulative percentages of errors
were constructed in order to better interpret the results. See Figures
A3, A5 and A7.
2.3.4 Conclusion
�4.S.rThis section will attempt to draw together those points of
the study which need to be emphasized. Reviewing Figure AS for
overall dimensions it shows (as the dotted lines
indicate) cumulative percentage values of errors at both the 90 and
the 50 percentile marks. Following the dotted lines over to the
curve and down, it is seen that 90% of the errors are less than
equal to
ft.
and 50% of the errors are less than or equal to
or
X Since the most interest is in small dimensions (2.0 ft. or
less), Figure A5 was constructed showing the errors of those
dimensions ranging from ito 2.00 ft. On this graph, 90% of the
TCS 35974-77
85
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
errors are less than or equal to
less than or equal to
and 50% of the errors are
...4&)- If only the dimensions less than or equal to one foot are
graphed, the data sample becomes so small that it is difficult to
make an analysis. Figure A7 of the cumulative percent of errors for
these short dimensions shows that 90% of the errors are less than or
equal to and 50% of the errors are less than or equal to
XIn conclusion, the results show that for dimensions less than
or equal to 2.00 ft., the probability of a measurement being within
of the true value is 90%, or 90 out of 100, and being within
of the true value 50%, or 50 out of 100. In a like manner,
the same statement can be said for the two other data sets. It must
be pointed out that the stated probabilities are applicable only for
the data sample (photographs) used here.
2.3.5 Summary
J,S74.11C.) Probably the most interesting and impacting quality of
the Tacana system is its ability to -see- small dimensions that
otherwise could not be seen on KH-8 imagery, and be able to
measure these dimensions accurately. Limiting the test to just one
target area reduced the data sample considerably. Limiting the
study to two good photographs further reduced the data set.
However, it must be emphasized that the exercise, being a feasibility
study, was very practical. The results can be assumed to be
indicative of those for an operation such as Tacana where image
acquisition has not been optimized to its full potential.
2.4 Interactive Digital Image Manipulation (IDIMS)
2.4.1 Introduction
),Wf Test imagery acquired by Tacana over the Washington
Navy Yard was manipulated on the IDIMS to illustrate the
capability of the system to improve smeared imagery for exploita-
tion. Because Tacana has no exposure control, imagery ethibiting
heavy shadows was also selected for analysis.
(1,,,,1 /A-117Q The IDIMS is located in Room 4N 814, Building
213, and is utilized to extract additional intelligence information
from problem imagery for photointerpreters and analysts. The
IDIMS consists of a Hewlett-Packard 3000 CX computer with
peripherals (i.e., three magnetic tape drives, one line printer, one
card reader, four analyst terminals) and two COMTAL TV display
monitors. A DICOMED, Inc., image recorder is capable of produc-
TCS 35974-77
86
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
ing high quality hard copy results (black/white or color) on film or
Polaroid paper. Approximately 200 image manipulation processing
functions ranging from simple manipulation (e.g., contrast stretch-
ing) to complex manipulation (e.g., Fourier analysis and filtering)
can be applied to digital imagery. The target area on a film
transparency is digitized on a PDS microdensitometer and then
manipulated on the IDIMS to achieve the desired results.
(U) Upon selection of the frames for analysis, an area from
each was digitized on a PDS microdensitometer using an 8
micrometer sampling and step-over interval, and an 8 micrometer
scanning aperture. The digital array generated on the PDS was 512
samples x 512 lines. This array covers a square area (4.096mm x
4.096mm) on the Tacana ON film. Figure A32 (a-d) illustrates the
effects of IDIMS on a smeared Tacana Navy Yard acquisition.
2.4.2 IDIMS Technique
The 512 x 512 microdensitometer scan was reduced by a
factor of two (i.e., making the image equivalent to a 16 micrometer
scan) to speed up processing in the Fourier transform domain. No
information was lost in the reduction process since this frame
contains much less fine detail than the other three frames and is
shown in Figure A32a. A Wiener filter/image motion correction
routine was first applied to the Fourier transform of the image with
little or no improvement.
j. A second Fourier transform technique, defined as "rooting"
and similar to applying a high pass filter to a Fourier transform was
applied to the magnitude (i.e., amplitude) of the Fourier transform.
This technique raises the magnitude to a power and is an alternate
means for correcting for moderate image motion and for enhancing
edge detail. Figure A32b is the result of the rooting technique and
Gaussian filter only. Figure A32c represents the effect of two
additional edge enhancement techniques (to A32b). The exponential
value of the function was then changed to 0.5 (i.e., similar to
applying a very high pass flter). The resulting image was then
inverse transformed and a 5 x 5 low pass convolution filter applied
to reduce the noise. This result is shown in Figure A32d. Much
sharper edge definition is apparent in this result than in the original
degraded image.
(U) Inverse filtering, i.e., dividing the Fourier transform by the
image motion equation (sin wx/wx) to remove the image motion,
was not attempted. Also, the phase portion of the Fourier transform
was not corrected by direct inverse filtering.
87
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
2.4.3 DICOMED Results
UO) The hard copy results listed above were generated
on Plus-X 120 film in the DICOMED high resolution mode. This
film transparency was then processed in the Versamat at a speed of
ten feet/minute. The DICOMED film copy is comparable to results
viewed on the IDIMS. An infinite number of results can be
displayed on the IDIMS using a trackball cursor which controls the
contrast/brightness upon completion of the more sophisticated
IDIMS routines.
2.4.4 Summary
IDIMS improved the smeared frames used in this evaluation.
Some loss in "enhancement- effect is noted in the intermediate
paper prints as PIs gain most benefit from the IDIMS directly from
the viewing screen.
,,51)" Possible future Tacana experiments, if deemed feasible,
should center on restoring smeared imagery using a number of
existing routines with varying parameters. The maximum a' pos-
teriori (MAP) method of restoring images should possibly be
investigated. The MAP algorithm processes small sections of the
image sequentially and pieces them together to create the restored
picture. The Office of Research and Development/CIA is currently
evaluating this technique
Research and application of the MAP restoration
technique to operational smeared photography in 1976 indicated
some improvement in image quality. Tacana imagery not degraded
by image motion or defocus can be manipulated with existing
IDIMS functions.
Section 3 Image Quality Analysis and Film Selection
3.1 Introduction
3.1.1 erVilm selection for the Tacana system was based upon
an examination of quality factors desired from the product which
were appropriate to the exploitation objectives of this study. The
quality goals were:
a. The film should have a film speed and latitude to record a
wide range of target ref lectances under variable light
conditions.
b. The camera/film combination should have a recording
potential of 5.08cm (2 in.) Ground Resolved Distance (GRD).
c. The prime film consideration will be black-and-white. Color
films will also be investigated for their applicability.
TCS 35974-77
88
Top
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
d. Other factors which affect mensuration, field handling,
camera/film compatibility, interpretability, etc., will be con-
sidered and maximized.
3.1.2 (U) The specific quality factors examined were:
a. Film Speed
b. Resolution
c. Exposure Range
d. Processing Flexibility
e. Base Thickness
f. Granularity
g. Field Application
h. Availability
3.1.3 (U) Final film recommendations were based upon a
weighted scoring which takes into account the relative importance
of each quality factor to produce an optimum image for exploitation
purposes. See Section 3.2.6 for scoring methodology. Table A6
illustrates the scoring technique and weighted quality factors used
for each film that was evaluated. The preceding factors and their
relative significance to this system product would not necessarily be
the same for other systems.
,(i?(For this study, there were two unusual constraints:
a. The camera is a fixed exposure type allowing no control for
film speed or varying light conditions.
b. The pigeon platform is unpredictably variable in the direc-
tional nature of his motion, flight path, and velocity. The
following sections will describe the techniques and results of
the evaluation of each quality factor used in the film
selection process.
3.1.4 System Parameters
(U) Table A2 describes those characteristics of the camera,
platform, target, and light conditions that were used for the analysis.
Where no data was available, estimates (noted) were made for
purposes of calculations.
3.2 Film/Quality Factors Evaluation
3.2.1 Film Speed/ Latitude
(U) Figure A8 represents the estimated exposure range we
would expect to record with this system for low altitude photogra-
89
TCS 35974-77 T ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF-
phy. Basic assumptions ror the calculations were taken from Table
A2. The exposure values were calculated from:
E=IT
where E=Exposure in meter candle seconds
I=Intensity in meter candles
T=Time in seconds
Image Illumination was calculated from:
IF. 0.64B
N2
where IF=Image illumination
B=Surface Brightness
N=f/Number
0.64=estimated lens transmission factor
(U) The film curves shown in Figure A8 represent three
significantly different effective aerial film speed (EAFS) materials.
They are typical for those films and do not represent an effect of
unique processing or handling.
(U) An analysis of this figure also provides the basis for an
estimate of the film type which would be required to obtain a
proper system exposure and dynamic range. It reflects the exposure
values (Log Exposure) for shadow and non-shadow areas of a target
given:
a. A fixed shutter speed (1/2200 sec) and aperture (f/2.7)
camera.
Table A2�System Parameters
Camera
Mod. I (MCW24) Mod. II
Lens type Minox
Aperture f/3.5
Focal Length 15mm 15mm
Film Size 16mm unperf. 16mm unperf.
Format 12.5 x 7.5mm 12.5 x 12.5mm
Shutter Speed .0007 sec. .00045 sec.
(1/1400) (1/2200)
Platform (avian)
Forward Velocity variable; 40 knots*
Pitch and Roll Rate variable; 40 knots*
Propulsion (wing beat) Rate 6/Hz
Flying Height variable; 30.48M (100 ft)*
Target (Nominal, Intelligence Type)
Contrast 1.6:1
Reflectance 12%
Lighting *(Sunlight,50�-60� Solar Elevation)
Direct Illumination 10 Foot-Candles
Open Shade 10 Foot-Candles
* Estimated for preliminary calculations.
1
1
1
1
1
TCS 35974-77
90
Top ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top ret RUFF
b. A target of esti ated 2% reflectance.
c. Various solar elevations.
(U) The range lines indicated by shadow and bright sun
represent the scene exposure range. The ideal film (in terms of
speed and latitude) would be represented by the straight line portion
of its characteristic curve falling between the exposure range lines.
By moving the left exposure range line to coincide with an
appropriate solar elevation, the required log E can be estimated for
various solar elevations for a clear (sunlighted) acquisition.
The above analysis and subsequent testing determined that
3400 series (3410) film was the best compromise of all the
considered quality factors. Eastman Kodak was also requested to
determine the availability of any new films which might be
compatible for the system. They indicated that a candidate film was
in the R&D stage and would be available in early 1977 for testing.
3410 material was then used for all system testing until arrival of
the experimental material (FE 6526) in June. It was evaluated and
found to be more compatible than the 3410 (see Section 3.2.6).
3.2.2 Resolution' GRD
,(*(A determination
potential was made
Missouri. The resolution test consisted of suspending the camera
from a balloon and photographing resolution targets of known
contrast with color and black-and-white films. The materials used
for baseline data were Eastman Kodak 3410 and 1414. They were
selected because initial calculations indicated that the 3410 material
had a film speed (EAFS) that would just satisfy the system exposure
needs, and the 1414 would provide a high resolution base line to
compare the impact of film resolution needs against exposure and
camera performance. Figure A33 illustrates the test range. Objects
in the scene were those readily available at the site. They were used
for subjective impressions of image quality and initial mensuration
estimates. The resolution targets were 100:1 contrast type (AF1959)
graduated in 6th root of 2 increments.
(U) Figure A9 shows the resolution data obtained from this test.
Each point represents the average of two to four resolution readings
(based upon the individual coverage) from seven frames. The
highest resolution readings were consistent with several prior
ground/bench resolution tests. Although this test was done only for
the Model I MCW-24 (Minox lens) camera, subsequent tests on the
Model II !ens) camera indicated that it is representative
of the camera/film performance. Also included in Figure A9 are the
film resolution specifications as supplied by the manufacturer. The
of the camera/film system resolution
91
TCS 35974-77 Top
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
right axis indicates the calculated GRD at 100 ft. altitude when the
resolution performance of the system is as noted on the left axis.
The bottom axis indicates target contrast. For example, a 50 c/mm
image taken at 100 ft. of a 2:1 contrast target could theoretically
produce a GRD of 0.8 inches on 3410 film.
J,S..)' Analysis of this balloon test data indicated that the camera
was the primary resolution determinate (or the 1414 resolution
would have been considerably higher). Frame-to-frame variability
was high. Nonetheless, because the camera was well focused and
checked out prior to the test and no malfunction occurred, it is
assumed that its variability is part of the system and, at worst, its
potential performance at 100 ft. altitude could provide (on the
average) better than 2 inches GRD (at 1.6:1 contrast) on 3410 series
film.
3.2.3 System Evaluation
4.1'53 Subsequent to the resolution and exposure evaluation of
the camera and film from the balloon platform, a test of the total
system (bird, camera, film) was undertaken over Andrews AF to
determine the effects of the avian platform motion dynamics on the
image quality. These tests were to combine a bird location and
training exercise with additional image quality and camera opti-
mization experiments. Also it would begin to give a better estimate
of the problems associated with operations over large cultured areas.
Prior to this, only operational flights over basically non-cultured
areas had been made and these had been difficult to evaluate from
an image quality point of view. Table A3 indicates the estimated
performance of the system on the Andrews tests with camera Model
I (MCW-24). Photoscientists evaluated 854 images from five test
flights. The quality determination was based upon a subjective
estimate of GRD. A three-level scale was used for the judgments:
(1) Better than 15.2cm (6") GRD (about NIIRS 8-9)
(2) From 15.2cm (6") to 76.2cm (30") GRD (about NIIRS 5-7)
(3) Worse than 76.2cm (30") GRD (about NIIRS 1-4)
X Examination of the data from Table A3 indicated that the
total system could perform with a potential quality (GRD) ap-
proaching study goals. This data also indicated that (1) performance
Table A3�System Performance Over Andrews AFB
GRD
Flikht No.
No. of Images
Better than 6-
6--30-
Worse than 30-
1
251
22%
37%
41%
2
235
15%
41%
44%
3
123
7%
34%
59%
4
117
5%
65%
29%
128
6%
43%
51%
92
TCS 35974-77 Top
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
decreased with succeeding flights, and (2) even at best (Flight #1)
the probability of a system acquisition approaching the GRD goal
was low (about 22%). These observations led to a camera examina-
tion which showed a shutter spring breakdown, and that a faster
shutter speed was needed to increase the number of high GRD
pictures. At that time the camera contractor was developing a
Model II camera unit utilizing a special design f/2.7 lens
(to replace the f/3.5 Minox lens) and a larger image format. The
larger lens aperture was initially proposed to allow a slower, higher
resolution film to be used with the system. Based upon the
preceding GRD/film performance test data and the Andrews GRD
evaluations, it was felt that the unpredictable motion characteristics
(altitude and velocity) of the avian platform was the greatest single
contributor to degraded imagery. Therefore, the decision was made
to increase the shutter speed to produce a higher percentage of
sharp (non-motion degraded) images. The following formula, using
known and estimated platform dynamics of roll, maximum velocity,
etc., determined that in excess of 1/2000 sec. was necessary to
assure more consistent, non-smeared imagery. The first Model II
prototype flight incorporating the improved shutter indicated this to
be so. This complete flight (85 images) was NIIRS rated by
NPIC/IEG PIs for confirmation and the data is shown in Section
2.1.
Required
Platform Flying Height
_
Shutter Speed (1000) (Required (Lens Focal (Platform
Resolution) Length) Velocity)
3.2.4 Processing and Other Film Characteristics
3.2.4.1 (U) Numerous combinations of films and developers
were tested to determine if some specialized process might optimize
some specific quality factor of a film to make it more effective for
this system. Developers used included: PUSH-POTA (NPIC formula-
tion), H and W, Kodak D-76, Kodak D-19, Rodinal, and Kodak
HC110. The primary process factors evaluated included film speed
manipulation, contrast control, and field (operational) utilization. All
films were not evaluated with all developers, but an effort was
made to determine if certain high resolution films could be made of
an appropriate speed and dynamic range without compromising
other factors of handling and image quality. Sensitometric data was
collected where possible (1414, 3410, FE6526) for evaluation.
(U) Following is a brief summary of the film/developer
examinations:
a. PUSH-POTA (NPIC formulation)
(UjiirtrO) Effectively increased the EAFS of all films that were
processed (1414, 3410, 5069) about 2 times. But 1414 (at an
TCS 35974-77
93
Top ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
T cret RUFF
EAFS of 15) was still too slow, 3410 did not need such a
magnitude of speed increase, and 5069 gave almost an equiva-
lent speed as 3410 but did not have an appropriate latitude.
Additionally, PUSH-POTA produced dichroic (chemical)
fog on the 1414 and had an extremely high fog density with
the 3410. Field use is possible but a 95�F processing tempera-
ture requirement and questionable storage life makes its han-
dling difficult.
b. Kodak D-76
(U) Worked well with 3410 and FE6526 films. Clean working,
stable, and readily available. It produced about a 50% speed
increase with a moderate loss of dynamic range.
c. Kodak D-19
(U) Worked well with 1414 but produced a very limited
exposure range product for this system. It is designed to add (with
1414) contrast to high altitude aerial acquisitions that are typically
of low overall scene contrast (haze effects, etc.). It did not increase
the 1414 EAFS enough to make it usable. Produced a high contrast
image with the 3400 with some speed benefit in the highlights.
d. RODINAL
(U) Produced acceptable imagery with 3410 and 410. Has
potential for further evaluation, but it is difficult to handle in the
field requiring very precise (syringe) application.
e. Kodak HC110
(U) Good results with FE6526 film. Readily available and clean
working. With a matrix of tested dilution rates and processing times,
it has the capability to alter film speed and contrast relative to the
camera exposure/target (scene) brightness ratio. This was used for
the 6526 film the last few weeks of the study and the results
obtained were sensitometrically and subjectively acceptable.
3.2.4.2 Base Thickness and Granularity
(U) For this system a 2.5 mil. base thickness film is most
appropriate in terms of film load (more exposures/flight) and
handling. Granularity of a film has an effect on mensuration and PI
exploitation. It is difficult to measure the total benefits of a finer
grained material with this system's resolution capability, but it is
enough to know that a lower granularity usually provides better
exploitation performance.
TCS 35974-77
94
Top ref
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
3.2.5 Color Films and Analytical Application
3.2.5.1 �,)1 The system restrictions of the fixed exposure
camera, variable lighting conditions, limited process manipulations
available (speed pushing, etc.) and inherent lower resolution capabil-
ities of color materials made selection limited. The films evaluated
were SO-255, 2445, 2448, Vericolor II, SO-397 and SO-131 (false
color infrared). All were positive types except 2445 and Vericolor II.
The final selection (EK 2445, Figure A31) was made based upon its
speed, resolution characteristics and wide (relatively) exposure lati-
tude. It is well suited for duplicate positive reproduction, paper
print enlargements, process manipulation and color correction.
.a.1.74115V) NPIC is presently contracted with Calspan Corpora-
tion to implement SCAT onto the Interactive Digital Image Manipu-
lation System (IDIMS). The technique is probably applicable to the
Tacana system but, as yet, has not been tested against negative color
materials.
*Smith, Turinetti, RADC's Research in Color Image Interpretation, Journal
of Applied Photographic Engineering Volume 3, Number 2, 1977.
Table A4�SCAT Materials Discrimination
TCS 35974-77
,,7ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
3.2.6 Film Selection
3.2.6.1(U) Table A5 summarizes the basic characteristics of the
films that were considered for this system.
(U) The films selected and illustrated in this report represent
materials that performed best overall as indicated by the weighted
quality/performance scoring criteria (Table A6).
(U) The performance value (1 to 10) indicates how well each
film performed in satisfying a given quality factor. The sum of all
quality factor scores (WF X PV) was added to provide a final total
value for each film. (See Table A7.)
Table A5�Film Characteristics
BLACK-and-WHITE
Speed
Film Type (EAFS)
Resolution
1000:1
(c/mm)
1.6:1
Granu-
larity
Thickness
(MILS)
3414/1414 (EK)
9-15
630
250
9
3.2
High Definition Aerial
FE6526 (EK)
80-100
350
150
10
2.9
Experimental, Poten-
tial to Replace 3400
Series
3410 (EK)
80-100
240
90
20
2.9
PAN-X Aerial
3401 (EK)
200
95
35
32
3.1
Plus-X Aerial
5069 (EK)
80-180
250
100
4.5
High Contrast Copy
SO-410 (EK)
100-200
250
100
6
4.5
Photo Microphoto-
chrome
VTE -80" (H&W)
80-100
160
70
4.5
Panchromatic
COLOR
SO-242/255
6-8
200
100
11
3.7/2.7
High Definition Aerial
(positive)
2448
32
80
40
12
4.8
Ektachrome MS (posi-
tive)
2445
100
80
40
13
4.9
Aero Color (negative)
Vericolor II
100
70
30
�
5
Commercial (negative)
SO-397
64
80
40
13
4.9
Ektachrome EF (posi-
tive)
SO-131
40
63
32
17
4.8
Aerochrome Infrared
(2443 type)
Table A6�Film Scoring Factors
Film Quality
Factor
Weighting
Factor (WF)
Performance Score
Value (PV) (WF) X (PV)
Film Speed
5.5
(1-10)
Resolution
5
Exposure Range (Latitude)
4
Processing Flexibility
3
Base Thickness
3
Granularity
2.5
Field Application
2.5
Availability
1
TCS 35974-77
)545.4,
Top ret
1
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
cret RUFF
Table A7�Film Scores
Black-and-White
Color
Film
Score
Film
Score
*FE6526
204
2445
163
3410
194.5
SO-397
159.5
SO-242/255
148.5
5069
188
2448
141.5
3411
182
VERICOLOR II
138
1414
179.5
SO-131
100
VTE "80"
177.5
* This film is presently experimental. Eastman Kodak indicates that it will replace 3400 series Aerial
Reconnaissance Films in the near future.
3.2.6.2 Recommended Films and Processes
Black-and-White: Film�FE6526 (Eastman Kodak) Film-3410 (Eastman Kodak)
Developer�HC110, Dilution D Developer�D-76, 6 1/2 minutes 21�C (70�F)
Six minutes 21�C (70�F) Constant Agitation
Constant Agitation
Color: Film 2445�Eastman Kodak Aerocolor Negative
Standard Color Process as recommended by the manufacturer.
3.2.7 General Camera Anomalies
(U.1�9-1430) The photographs in this report are used to describe
The exploitation quality of the system, as such they are so
referenced in preceding sections. They also show the effect of
various camera problems associated with its development. Directing
attention to these anomalies is only intended to indicate factual
difficulties encountered in assembling system performance data and
indicates the need for further reliability production efforts. It should
be pointed out that the camera is a commendably engineered unit
and these effects are common and expected in this type of research
effort and could be corrected in the routine modification and
upgrading of the camera units for field use.
Demonstrate the effect of non-uniform sharpness due to prob-
able misalignment of the film plane with the optical axis. Decentra-
tion of an element in the optical system or an imperfect and/or
improperly positioned field corrector would produce a similar effect
(noted in the lower central portion of these photographs).
The off-center light area in these photographs represent an
effect of internal lens flare (stray light reflected to the film by
improperly baffled optical components). This fogging of the film has
the effect of producing a lower contrast and resolution image. It
was present in most cameras in varying degrees.
Represents the "barrel" lens distortion effect present in the
(Model II camera) lens (curved portion of waterline edge
at bottom of frame). This type of distortion reduces towards the
center of the image format.
97
TCS 35974-77 cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
ToS 35974-77
98
ecret
SECRET
andle via
TALENT-KEYHOLE
Channels
TOP
1
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
CUMULATIVE FREQUENCY
CUMULATIVE % OF ERRORS <
Top cret RUFF
0.5
06
07
0.8 0.9 1
11
11.2
13 FEET
1 1 I I 1 1
0.0 2.0 4.0 6.0 8.0 10.0
1 1.0 25 15.0 20.0 30.0
35.0
40.0 CM
ABSOLUTE ERROR
FIGURE A2. Histogram of Cumulative Frequency of Error in Measurements of 53 Dimensions
(.25 Feet - 37.5 Feet)
100
75
50
25
0.0
0.1
0.2
0.3
0.4
0.5
I
I
I
I
I
0
2.0
4.0
6.0
8.0
10.0
15.0
UNCLASSIFIED
0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 FEET
20.0 25.0 30.0 35.0 40.0 CM
ERROR
FIGURE A3. Graph of Cumulative Percentage Error for Measurements of 53 Dimensions
ITCS 85974-77
Approved for Release: 2019/07/31 C06527327
UNCLASSIFIED
Approved for Release: 2019/07/31 C06527327
CUMULATIVE FREQUENCY
10-
9
8 �
7 �
6 �
5 �
4 �
3 �
2
1-

0
00
ecret RUFF
.10
.20 .30
0.0 2.0 4.0 6.0
8.0
10.0
.40 FEET
CM
ABSOLUTE ERROR
FIGURE A4. Histogram of Cumulative Frequency of Error for
22 Dimensions of 2 Feet or Less
UNCLASSIFIED
TCS 35974-77
100
Top ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top ret RUFF
100 �
0.00 05 0.1 0.15
0.2 0.25 0.3 FEET
1-- I I i I 1 I 1 I 1 I
0.0 2.0 4.0 6.0 8.0 10.0 CM
ERROR
FIGURE A5. Graph of Cumulative Percentage Error for
Measurements of 2 Feet or Less
TCS 35974-77
101
To ecret
UNCLASSIFIED
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
CUMULATIVE FREQUENCY
00
0.10
0.20
0.30 FEET
0.0 2.0 4.0 6.0
ABSOLUTE ERROR
8.0 10.0 CM
FIGURE A6. Histogram of Cumulative Frequency of Error
for 14 Dimensions of 1 Foot or Less
UNCLASSIFIED
TCS 35974-77
102
To ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ret RUFF
VI
CUMULATIVE %OF ERRORS
100 �
50
0 00
0.05 0.10
0.15 0.20 FEET
0.0
2.0
4.0
ERROR
FIGURE A7. Graph of Cumulative Percentage Error
for Measurements of 1 Foot or Less
6.0 CM
UNCLASSIFIED
TCS 35974-77
103
To ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Solar Elevation(Deg)
7 10 15 30 50
co
0
Scene Brightness
Range
(12% Reflectance)
-4(
3401
(EAFS 150-300)
3400/FE 6526
(EAFS 60-80)
1414
(EAFS 6-15)
Log Exposure (MCS)
FIGURE A8. Graph of Predicted Target Brightness/Exposure Range and Film Estimates
ADMINISTRATIVE AL USE ONLY
TCS 35974-77
10
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
600
500
400
2
S 300
co
a)
200
100
50
*Manufacturers Resolution Data
**Camera/Film Performance
(Average of 7 Test Simulations)
FE 6526*
1414**
---- 3410**
3410*
........ ...-����
......
........ .........
....... .������-
......... .........
../ ......-
........"
------ ............
�������
������
2445*
GRD (at 100 Ft)
.75
2:1 10:1 100:1
Contrast
FIGURE A9. Graph of the Camera/Film Performance
Evaluation and GRD Estimates
Co
a)
1.0 -=
2.0
1000:1
SEC
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A10. Washington Navy Yard-3/17/77
Rated NIIRS 9
TCS 35974-77
106
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
T Secret RUFF
FIGURE All. Washington Navy Yard-3/17/77
Rated NIIRS 6
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
........loirFF
FIGURE Al2. Washington Navy Yard-3/17/77
Rated NIIRS 7
108
To ecret
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top ret RUFF
II
II
II
II
TCS 35974-77
FIGURE A13. Washington Navy Yard-3/17/77
Rated NIIRS 7
109
Top ret
SECRET
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A14. Washington Navy Yard-3/17/77
Rated NIIRS 8
S RET
NOFORN
TCS 35974-77
110
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
'FIGURE A15. Washington Navy Yard-3/17/77
Rated NIIRS 8
TCS 35974-77
111
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
T ecret RUFF
FIGURE A16. Washington Navy Yard-3/17/77
Rated NIIRS 9
SECRET
NOFORN
1
TCS 35974-77
112
To. -cret
Approved for Release: 2019/07/31 C06527327 .
v.
Approved for Release: 2019/07/31 C06527327
Top et RUFF
FIGURE A17. Washington Navy Yard-3/17/77
Rated NIIRS 9
SECRET
NOFORN
TCS 35974-77
113
To. -cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGUREA18. Washington Navy Yard-7/7/77
Maximum Observed Flying Height
300 Ft. (91 M.)
NOFORN
TCS 35974-77
114
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
FIGURE A19. Washington Navy Yard-7/7/77
Maximum Observed Flying Height
300 Ft. (91 M.)
G T
NOFORN
TCS 35974-77
115
cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
FIGURE A20. Washington Navy Yard-7/7/77
Maximum Observed Flying Height
300 Ft. (91 M.)
SE T
TCS 85974-77
116
To cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
II
II
11
11
It
FIGURE A21. Washington Navy Yard-7/7/77
Adjacent Frames (with Figure A22)
Showing the Effect of Platform
Dynamics on Image Quality
RET
NOFORN
TCS 35974-77
117
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Secret RUFF
FIGURE A22. Washington Navy Yard-7/7/77
Adjacent Frames (with Figure A21)
Showing the Effect of Platform
Dynamics on Image Quality
NOFORN
1
TCS 35974-77
118
To ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
ecret RUFF
FIGURE A23. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
119
Top ecret
RET
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
ecret RUFF
FIGURE A24. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
120
ecret
, T
e�E
OFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
TCS 35974-77
FIGURE A25. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
121
Top ret
....."Clifr��NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A26. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
...."CitErNOFORN
TCS 35974-77
12s,t2e,,,,��
To cret
Approved for Release: 2019/07/31 C06527327
-
Approved for Release: 2019/07/31 C06527327
FIGURE A27. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
5,641rE.<
NOFORN
35974-77
123
T ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
T ecret RUFF
FIGURE A28. Navy Yard Display Area
Not Mensurated
Stit
NOFOg
TCS 35974-77
124
To ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A29. Ground Views of Tabs Missile
at the Washington Navy Yard
UNCLASSIFIED
TCS 35974-77
125
To ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
TCS 35974-77
FIGURE A30. KH-8 Coverage of the Navy Yard
Target Display Area
126
Top ret
TO
3.3(
ECRET RUFF
NOFORN
h)(2)
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
FIGURE A31. The Washington Navy Yard-7/8/77
soFoRN
TCS 35974-77
127
To ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 006527327
Top ret RUFF
TCS 35974-77
32(a). Original Scene Micro-d Scan
32(d). Rooting (Exp 0.5) 5x5 Convolution, TIC
32(b). Rooting (Exp 0.7) Gaussian Filter 32 (c). Rooting (Exp 0.7) Gaussian Filter
Edge Enhancement
FIGURE A32. IDIMS Effect on Smeared Imagery
128
ecret
NOFORN
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 C06527327
T ecret RUFF
TCS 35974-77
FIGURE A33. Camera/Film Analysis Test Imagery
(From the Balloon Platform)
129
To. --cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top et RUFF
APPENDIX B
CAMERA DETAIL AND SPECIAL DEVICES
This section gives some additional detail on avian cameras
including results of some preliminary low temperature tests and a
quote from Mallory on a lithium battery for use in extremely cold
climates. Also discussed are the camera timer box, the camera
harness, the small avian DF transmitter, and a Skinner box
experiment for investigating the visual response of pigeons in the
infrared. As a result of the success of the Skinner box experiment,
an IR strobe was modified for use as a remote signaling device for
calling the birds into the loft.
Camera
87 Figures B1 and B2 are assembly drawings for the MCW-24
camera and the new camera. Low temperature tests were
conducted on each camera. It should be kept in mind, however, that
see OtTAILS
Loo
TCS 35974-77
FIGURE Bl. MCW-24 Camera Mechanism Assembly
131
Top cret
.OED(EO0,-r)VOSI4UTItI3.
. AIO
F50,0 ED
oerzLADeT ass',-.
mo...4 /1.4 El!, 5.4oTTEO.
rOO. S.IoI4I5 0.16, /155 Y.
UNCLASSIFIED
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
T/ 24T DUPLE.(
9T / 31 T DUPL.
ST 00702 Plf...110KI
&sis.rz PLA7 - f2A E
OTOR
2E412 LEIJS I-I DU S 1,1
SPI211.16 DETEKIT
SPIZIK14 STOP
COIL SPT21k1Gi -
SW LITTER �
ia.T/24 7 DUPLEX (2EF)
72T FACE. GEAR
(51-1UTTEf2 4AU5IN.14)_
82 T SPOOL Df2IVE G EA
DRIVE- 12 It.1
f6ATTEt2Y O(----
..�;.-FRO LII NI KIT LEW 140 S
TA VE � UP SPCxSL
FIGURE B2.
Camera Mechanism Assembly
ret RUFF
I
31T
IDLE f2 24T DLER
�/- IC0 T D LE 12
/ 31 T DUPLEX
FLATTENER7
UNCLASSIFIED
the MCW-24 was flown 12 times over Andrews Air Force Base
during one of the coldest winters on record; typical temperatures
were in the low twenties (�F).
JeFil The first low temperature test was with the MCW-24
camera, using two Mallory 10L14B silver cells. The unit was placed
in a ( � 5�F) freezer with a three-minute delay in the timing unit. A
thermocouple attached to the side measured camera case tempera-
ture. After three minutes, the camera started (case temperature
52�F) and ran for an additional four minutes (case temperature
40�F) at which time it stopped.
.(*). A second test was run under the same conditions, but with a
30 mph wind blowing over the unit. Here, the case temperature was
(35�F) after three minutes (when the camera started) and was
(23�F) after five minutes (when the camera stopped).
�(&)--In both the above experiments, the camera did not pull a
full load of film through before stopping. It is apparent that the
(107�F) body temperature of the bird kept the camera sufficiently
132
TCS 35974-77 To. -cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecret RUFF
warm to work properly during the flights over Andrews Air Force
Base.
An additional temperature test was performed on ths MCW-
24 using a constant three volt power supply. Here, the camera
stopped when the case temperature reached (32�F) and the current
drain was approximately double that at normal temperatures. The
10L14B silver cells were tested at low temperatures through a 39
ohm resistor. The voltage varied from 3.2 V (60�F) to 2.8 V (32�F)
to 2.0 V (0�F).
The camera was tested (no timer) with a constant
3-volt supply and continued to operate down to ( � 1�F). The
current drain increased from 36 ma (28�F) to 92 ma ( � 1�F). This,
however, was after considerable attentign had been given to
selecting a low temperature gear lubricant.
(U) Three factors contribute to the low-temperature per-
formance of the camera:
1. The current drain of the motor as a function of load and
voltage.
2. The viscosity of the gear lubricant at low temperatures.
3. The performance of the battery at low temperature.
(U) A 6-volt motor was found which had considerably better
efficiency than the current 3-volt motor. Tests with the 6-volt motor
gave 18 to 22 ma at (20�F) and 45 ma at zero. However, once a
suitable low-temperature lubricant is found, the critical item in the
systems is battery performance.
It is known that lithium cells have twice the voltage as
silver cells with virtually no performance degradation at low
temperature. Therefore, if lithium cells could be made in the same
size containers as the present 10L14B silver cells, the high efficiency
6-volt motor could be used to compose a system with adequate
performance at zero degrees Fahrenheit (or lower). This problem
was given to Mallory with the following response:
1. Size. Same as the 10L14B silver cell.
2. Weight. 2.6 grams/cell (same as 10L14B).
3. Current Drain. 100 ma at 2.6 v/cell for 15 minutes.
4. Cost. $180/cell recurring (in lots of 50), and $12,000
nonrecurring.
5. Availability. Four to six months.
6. Shelf-life. One to two years.
133
To ecret
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
Camera Time Box
The camera timer box is used to program the camera timer
electronics, which contains two E-cells for providing the -delay- and
-run- times. Figure B3 is a schematic, and Figure B4 is a
photograph of the control panel and the three timer leads. Figure
B5 is a picture of the camera electronics showing the positions for
connecting the timer leads. The following is a list of steps for using
the timer box:
1. With S-1 on and S-2 in -volts- position, the condition of
the batteries in the box can be determined. It should be
three volts, read on lower scale above the meter.
2. With S-1 on, S-2 in -milliamp- position, and S-6 in "open"
position, the camera should not run. If it does, it is because
there is accumulated time (charge) in the E cells of the
camera timing circuit. This should be allowed to run out
before proceeding further, preferably with the camera not
attached to the E cells, so as to save wear and tear on the
shutter.
3. With S-1 on, S-2 in -milliamp- position, and S-6 in "short"
position, the camera should start running and continue
running until S-6 switch is placed in -open- position.
4. To put run or delay time in camera, S-1 must be on, S-2
must be in -microamp- position, and S-3 must be in "time"
Run
S3-A Red
SHORT
2-A
FIGURE B3. Timer Box Schematic
White
Black
UNCLASSIFIED
TCS 35974-77
134
Top ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE 84. Timer Box Controls
UNCLASSIFIED
position. Further, if this is for an actual flight, batteries
must be in camera and a tab must be between ground and
contact points leading to batteries on camera.
5. Putting Delay Time in Camera: With timer set as
described in 4 above, S-4 switch is put in "delay" position
for amount of time desired. The meter should read
approximately 175 microamps (upper scale above meter)
while delay time is being put in. At the end of that period,
S-4 switch should be put back in "off" position.
6. Putting Run Time in Camera: With timer set as described
in 5 above, and with S-4 switch in "off" position, turn S-5
switch to "run" position for desired length of time. Meter
should read approximately 60 microamps (upper scale
above meter) while time is being put in. At the end of ,that
period, switch S-5 to "off" position.
7. To check delay and run time, disconnect three leads from
the camera and pull tab. Camera should delay running
until delay time is over, then run the amount of time
programmed.
TCS 35974-77
135
Top et
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Secret RUFF
Hook Black Lead Here
FIGURE 85.
amera Electronic Timer
Hook White Lead Here
8. To heal E cells, first disconnect the E cell portion from the
rest of the camera. Then with S-1 on and S-2 in -milliamp-
position, leave it for a period of time, first in "open"
position for S-6, then in -short- position for S-6.
9. After camera is loaded with film and prior to putting in
time for actual flight, five seconds of delay time and five
seconds of run time should be put in to test if camera is
functioning properly.
10. Unloaded, camera should be pulling around 75-85 milli-
amps when running. Loaded, it should be pulling around
100-125 milliamps when running.
Camera Harness
Figure B6 shows a bottom- and top-view sketch of the
harness. Several types of material were used with varying degrees of
success. The weight should not exceed 4.5 grams and a soft leather,
such as suede, is preferred which does not stretch with use or after
ET
TCS 35974-77
136
To ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
becoming wet. A heavy chammy was tried (the long axis of the
harness must be aligned � with the direction that does not stretch
much), but it must be prestretched prior to cutting out the harness
and is prone to stretching with excess use.
frir) The harness is connected (and adjusted) by use of Velcro
tabs on each end. The right side in Figure B6 winds up on top,
pointing aft, so that aerodynamic pressure does not peel the tabs
apart. On at least one occasion, the bird reached back and peeled
the tabs apart with its beak. For safety, the -tabs should be tied
together with lacing cord after the final adjustment. Since the
harness will loosen after the bird preens and works it into its
feathers (it becomes almost invisible from the bottom), the final
adjustment should be made in about ten to twenty minutes after
first being put on the bird. A light blue or gray� colored material
will blend in well with the underside of the bird. The cameras were
painted with a dull gray -automotive primer- which worked well.
A The Velcro camera tabs were used for rapid and convenient
attachment and removal of the cameras and weights from the
harness. However, this increased the total weight considerably and
was responsible for the loss of at least one camera. The first designs
of the MCW-24 cameras used small metal clips. Though this was
less convenient, it weighed less and was far safer.
,(e) A loose harness will cause the camera to bang against the
breast bone of the bird (a severe source of irritation) and result in
excessive photographic blurring due to high-angular rates. If the
Bottom View
TCS 35974-77
Camera Tabs
Tail
FIGURE B6. Camera Harness
137
Top Secre
Harness Tab
UNCLASSIFIED
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
cret RUFF
bird has a deep keel (breast bone), a small piece of foam should be
wedged between the harness and the belly. This will act as a shock
absorber for the vertical acceleration due to flapping. The bird will
move up and down about an inch, six times a second.
>#)' One size harness will not fit every bird. At the beginning of
the project, a large- and medium-size harness was used. As the birds
lost body fat and put on muscle, a third small-size harness was
necessary. It was found convenient to put the bird's number on its
harness. It is important to remember that the camera pull tab in
Figure B5 goes forward.
Avian Transmitter
431 Two types of avian transmitters (SM1 and SB2) were
obtained from:
Figure 12 shows the SB2 (with batteries, antenna, and ground plane
wire) attached to the harness. The SM1 weighs less than a gram (no
battery, antenna or ground plane wire) and can be detected
(ground-to-ground) at about one mile range; the typical life is over
one month. The SB2 weighs about 6 grams and can be detected at
several miles (ground-to-ground) with a life of about 10 days.
frFs). The SB2 was used with every camera flight at the
beginning of the project and was responsible for the recovery of one
camera. However, it was noticed that the birds' performance with
the camera was not as good as with the MCW-24 or the
simulated camera weights. On several occasions, the birds would go
down before returning home and pull the antenna and ground wires
off. Also, it is felt that the combination of the heavier
camera and the transmitter weight affected performance. Toward
the end of the program, no transmitters were used and performance
improved, though still less than with the MCW-24 or weight. Both
irritation (wires or floppy transmitter) and excess weight degrade
performance. If a transmitter must be used, the lighter the better
for performance.
Skinner Box Experiment
fe,Frr The loft in Missouri was managed primarily by
with a degree
in psychology, was familiar with much of the classic -Skinner Box
Experiments- with pigeons, had access to most of the scientific
reports, and had several professional acquaintances at the State
TCS 35974-77
138
Top cret
1
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
University at Roanoke. Her ultimate goal was to train the pigeons
(on the loft or in the air) to respond to an infrared flashing light by
quickly coming into the loft. Her experiments progressed in several
stages listed below:
I. There are three -light buttons' in the Skinner box which can
be illuminated with red, white, and green light. These
buttons close a contact switch when pecked by the birds. A
small computer was programmed to randomly illuminate one
of the buttons with red light and to feed the bird when the
red button was pecked. Figure B7 shows the bird about to
peck the red button, and Figure B8 shows the bird receiving
its food reward. The computer .also recorded data pertinent
to the rate of learning and differences between individual
birds. The TV camera in these two figures was used for
remote observation of behavior.
2. Once all the birds were trained to the visible red light, it
was replaced with an invisible (to humans) infrared light
with no detectable change in learned behavior. The wave
length was about 0.9 microns.
3. In the next experiment, an IR flashing light was installed
over the feeder in the pigeon loft. This was a commercial IR
TCS 35974-77
FIGURE B7. Bird Responding in Skinner Box
139
To ecret
UNCLASSIFIED
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE B8. Bird Receiving Food Reward
UNCLASSIFIED
camera strobe (see Figure B9) fitted with a two-second
strobe timer. The purpose was to form an association
between the flashing light and dinner time. This concept
worked well with the -dinner music- tape used by all the
project lofts (also developed by
4. The last phase was to place a flashing light outside the loft
as a cue to the birds that food was being served. The project
terminated before this phase was completed, but it appeared
that some birds did make the transfer in cue while others
did not. A pair of squabs (very young birds), raised during
experiment three, did come into the loft in response to the
flashing light after they had returned from their first flight
and looked as though they intended to spend the night on
the roof.
Whether it is a young bird out for the first time or a
relocated bird at its new home, the outside of the loft is strange and
unfamiliar compared to the inside. The IR flashing light could be
used as a signal of home. A similar method was used by the U.S.
Army. A particular colored symbol was used on top of each of the
small mobile lofts (which were moved as much as 20 miles a day)
so that the birds could not only find home, but distinguish one loft
TCS 35974-77
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
from another. This method, or some other visible cue, could aid
greatly in relocation. It was also noticed that there was some
correlation between the rate of learning in the Skinner box and the
flight performance in the field. Since the learning and recordkeep-
ing in the Skinner box is automated, a large number of birds could
be -graded- in a fairly short time.
FIGURE B9. Infrared Strobe and Timer
UNCLASSIFIED
TCS 35974-77
141
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
APPENDIX C
CARE AND HANDLING OF HOMING PIGEONS
Sanitation and Disease Prevention
.ker One of the most important aspects of this project has been
guaranteeing the health of the pigeons involved. This effort has not
been a total success because initially we were operating out of a
local existing loft and did not have complete control over our
pigeons. Consequently, during the project, we encountered three of
the most common pigeon diseases: Canker, Pigeon Pox, and Round-
worms. We lost two birds to Canker but survived the Pox and
Roundworms with no casualties; the birds were out of form for
several days. At one point during the project, Paratyphoid was
present in the loft, but our birds were not affected. A list showing
the symptoms and recommended treatment for these four diseases,
plus another very common pigeon disease, Coccidiosis, is given at
the end of this section. With complete control over our birds and
their environment, most if not all of these diseases could almost
certainly have been prevented. Over 90 percent of most pigeon
diseases can be prevented outright through proper methods of
sanitation and care.
I.
TCS
.4�1. Following are some of the most basic tenants of sanitation
and disease prevention:
1. The pigeon loft should have good ventilation with no drafts.
This ensures a constant source of fresh, clean air and helps
to keep the floor dry, which is imperative for pigeons to
thrive.
2. It is essential that pigeons receive fresh water at least once,
and preferably twice, a day. Water should be placed in the
loft so there is no possibility of contamination by pigeon
faeces, which contain the bacteria for most common pigeon
diseases.
3. Pigeons need fresh, clean grain that is fed in the proper
quantity (overfeeding leads to overweight, out-of-condition
birds), and that cannot be contaminated by the birds' feet,
which come in frequent contact with faeces. (Feeders can be
built which almost entirely eliminate the possibility of
pigeons walking in the grain the eat).
35974-77
143
Top -cref
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
4. Fresh grit should be provided daily, though this is not
mandatory. Pigeons use grit to digest their food. It is also a
source of vital minetals that the birds can get in no other
way. IfPgrit is kept before the birds, it should be checked
periodically to be sure it has not turned rancid.
5. The loft must be kept clean and should be scraped daily,
which takes only a few minutes.
6. Common body insects must be controlled. This is imperative.
Birds afflicted with lice and mites cannot perform properly.
If they do not become anemic, they will be worn out from
the constant drain of pecking at the source of irritation.
Affected birds can be heard stomping their feet at night in
an effort to rid themselves of the pests. Pigeons have been
known to desert their eggs rather than sit in an infested
nestbox. Virtually all body pests can be eliminated com-
pletely by the application of roost paint in the loft once a
month and by hanging a Vapona bar (Shell Pest Strip) in the
loft. (It should be replaced every three months.) The above
six points are basic to proper loft management. In the
project loft, constructed for kit 2F, these points were
adhered to and not one of the diseases present in the first
loft appeared. These and other important points of loft
management are treated in greater detail in Chapter 15 of
Dr. Leon F. Whitney's book, Reference 9. Several books and
periodicals are given in the list of references which contain
useful material on the care of homing pigeons.
XIn addition to basic sanitation methods, some fanciers have
established a system of disease prevention by treating their birds
quarterly for Canker, Coccidiosis, Paratyphoid, and worms. This
type of treatment is the same as treating for the disease itself as
discussed later. These and other supplies can be purchased from the
following pigeon supply companies:
1. Racing Pigeon Bulletin
94-G Compark Road
Centerville, Ohio
45459
2. Foy's Pigeon Supplies
Box 166
Golden Valley, Minn.
55427
3. Charles Siegel and Son
1011 E. Middle Street
South Elgin, Illinois
60177
4. C. A. Hammer Company
1512 S. 34th Street
Milwaukee, Wisconsin
53215
TCS 35974-77
144
Top et
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
lo�S�ef RUFF
(U) A very useful plastic-colored marking band, that cannot be
obtained from the above supply houses, was used extensively in the
project. It can be ordered from:
Boddy-Ridewood
41 Aberdeen Walk
Scarborough Yorks
United Kingdom
Factors Affecting the Performance of Project Birds
1. �,K The Moult. Pigeons renew all of their feathers once a
year. The moult is continuous from the daily moulting of the fluff
feathers until the heavy moult in the fall of the year. It takes at
least six months to moult the entire body of feathers. The moult
should not affect the performance of pigeons over short distances,
with the possible exception of the #10 primary feather�the last
and longest feather on each wing. It can be very painful for a
pigeon to fly when this feather is coming in.
fee) At one point during the project, one of our best birds had
three or four of the primary feathers on one wing broken off, and
we were unable to fly her for several months. These were pulled
out over the course of several weeks, except for the #10 (which
should never be pulled). After the bird gmiv new feathers in their
place, we resumed flying her.
2. Laying Hens. A hen about to lay should not be sent to
any distance. More than likely she will stay someplace along the
route home for several days until she has laid both eggs. After
laying the second egg, she will desert them and come home when
her cock does not relieve her at the time when he is supposed to
come on the eggs. The cock in a mated pair will begin to drive the
hen to the nest about five days prior to laying (this is called a
"driving cock"). The hen can be successfully flown up to two to
three days before laying. The hen can be successfully flown to short
distances within one or two days after laying. Most sitting hens are
highly motivated to return home at some point while they are
sitting on eggs�usually about 14 days�and again when the eggs
begin to hatch.
3. (54illaising Young. Both hens and cocks usually fly well to
their young, and' their performance under this condition will usually
meet or exceed normal expectations.
4. SO Driving Cocks. A driving cock should not be flown. He
may follow the first hen he sees to wherever she is headed.
TCS 35974-77
145
Top ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
5. fel Hunger. Virtually all pigeons will respond well to
motivation through a restricted diet and will hurry home and trap
quickly once they learn food awaits them there.
6. 0 Launch Site Familiarity. The more familiar with the
launch site, the more uniform the departure for home will become.
After 5-10 tosses, a bird will depart in a fairly predictable manner,
although not nscessarily covering the same ground as before. The
danger of launch site familiarity is that as the bird tires of the
harness and weight routine, he will be more inclined to land in the
increasingly familiar area of the release site, unless proper motiva-
tion, (i.e., hunger) is maintained.
7. ,ce) Wind. Winds under 10 mph do not seem to affect bird
performance. Pigeons tend to tack into the wind, however, when it
is over 10 mph. This can be advantageous where winds are coming
from the target area, resulting in a slower ground speed.
8. ker Single Tossing. Pigeons are very gregarious and when
single-tossed tentl to remain in the area of the release waiting for
other birds with which to fly home. This could be advantageous if
lingering in the area were desirable. Double tossing was used with
good success in the project to get the birds to move out quickly for
home.
9. .(.�0 Harness and Payload Devices. When initially fitted with
a harness inside the loft, most pigeons in the project either went
into wild gyrations trying to get it off or sulked quietly in a corner.
After the harness had been on a day, all birds were functioning
normally around the loft, except for occasional picking at the
harness with their beaks. Later, when a 40 gram weight was
attached to the bottom of the harness, it did not seem to have any
affect on their performance on the ground, other than a high
-goose-stepping- walk as their feet brushed the sides of the weight.
Most birds even successfully sat on their eggs with the weight on
without breaking them. The weight definitely slowed them down on
the wing, however. The harness alone seemed to have little effect on
their desire to fly once the birds became used to it. Some harness
and weight training can be done in the loft.
Relocation
0 Because of its strong urge to -home,- a good homing
pigeon will make every effort to do so when released in strange
territory. Nevertheless, for various reasons fanciers periodically
attempt to relocate homing pigeons to a new loft at some distance
from the old. It has been done successfully many times, but the
danger of losing the bird is high. Dr. W. E. Barker in his book,
1
TCS 35974-77
146
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release 2019/07/31 C06527327
Top ret RUFF
Pigeon Racing, (see Reference 6) mentions the following points as
essential in relocating pigeons:
1. The better the bird, the easier it will be to settle to a new
home.
2. It is generally useless to attempt relocating birds which have
bred elsewhere to a new home until they have been allowed
time to breed again amidst their new surroundings.
3. Birds to be relocated should be given the opportunity of
becoming acquainted with the outside appearance of the
new loft and as much as possible with the new
neighborhood.
4. Birds in the process of being relocated must be handled with
great care and gentleness. They should never be startled
when allowed out for the first time.
5. Do not attempt to relocate hens between eggs or cocks while
driving. Rather, give them their liberty while sitting and
allow them to find their own way out of the loft without
interference.
6. Never attempt to relocate birds that have small youngsters in
the nest. They will almost invariably return to their original
home.
7. So long as relocated pigeons are allowed to remain undis-
turbed in their new quarters, they frequently exhibit little
tendency to return to their old surroundings. Once they are
disturbed or unsettled in any way, however, for instance
being sent to races, the desire to return to their old home is
apt to reassert itself, and they may return to their original
home.
With relationship to this project, the difficulty of relocating
pigeons is especially evident from point 7 above, inasmuch as
relocated project birds were -disturbed- when fitted with a harness
and weight and released in unfamiliar surroundings.
,er Following are the methods used in this project to assist in
relocating pigeons:
1. All birds were kept in confinement a minimum of three
weeks.
2. Birds were not allowed to see the outside of their new loft
prior to release because of operational considerations.
3. Some birds were allowed to raise a round of youngsters
before being released in an effort to cement the bond to
TCS 35974-77
147
Top et
Approved for Release: 2019/07/31 C06527327
Approved for Release 2019/07/31 C06527327
cret RUFF
their new home. However, most were not on eggs at the
time of release.
4. Some birds had one wing -hobbled" with a rubber band the
first few times out until they became familiar with their
immediate surroundings.
5. Three birds were placed on the -Widowhood" system of
flying (to be explained later) in an effort to relocate them.
K Most of the birds that we attempted to relocate were
eventually lost, most likely for the following reasons:
1. They were not able to familiariZe themselves with the
outside of their new loft and, consequently, many were lost
almost immediately upon release.
2. Those birds kept in confinement the shortest period of time
(three weeks) were the most likely to be lost. In general, the
longer the birds were kept in the new loft, the better the
chance of successful relocation, although still poor.
3. Many of those birds that survived the initial release were
subsequently lost when fitted with harnesses and weights.
Generally, they did not leave while -in harness" but did so
on training flights while not harnessed or without weights. In
other words, they were disturbed and left at the first
opportunity when they were not encumbered with a harness
or weight. In terms of initial success in the relocation
process, the best statistics were obtained with the three cocks
on the Widowhood System�three out of three. Nevertheless,
two of these birds were subsequently lost�one off the loft
just prior to a severe thunderstorm and the other, while
wearing a harness only, from the nine-mile training station
(again, see Barker's point #7).
4. A contributing factor in the loss of many birds was the
location and appearance of the new loft compared to the
old. In no case were the two similar. Moreover, in at least
one relocation series, the loft was so surrounded by trees that
most birds were lost as soon as they took flight. Hobbling at
this location gave better results.
The Widowhood System
1,)2`r The following is a brief and over simplified description of
the system:
1. Allow X number of mated pairs to pick their own nest boxes
in the loft (no other birds are allowed in the loft) and raise
TCS 35974-77
148
Top ret
1
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ret RUFF
one round of youngsters; two rounds is better if the birds are
yearlings, but it takes a long time (about two months).
2. After the birds go back down on eggs, allow them to sit for
about one week, then remove the hen and throw the eggs
away.
3. Hens should be kept in a place where the cocks can neither
see nor hear them. After separation, do not allow the cocks
to see their hens for at least a week, then take the cocks out
of the loft, lock the hens in half of the nest box, and leave
the other half open (so the cocks cannot get to the hens
when they are allowed in the loft). When the cocks are
allowed to enter the loft they will be in a high state of
excitement when they see their hens. Let the cocks in with
their hen but for no more than five minutes. Do not allow
the cocks to mate with their hen or the system will be
ruined. After five minutes remove the hen and put her back
in her loft.
4. After several weeks of doing this, the cocks will learn that
when they are taken from the loft, the hen will always be
waiting for them when they return. When the cocks are let
out for daily exercise, they will fly long and far searching
for their hen and will come into excellent physical shape.
When taken away for a race (or whatever) cocks will
speedily come home and trap in quickly, assuming the
system is working at its best. The disadvantage to the system
is that there are many variations, each quite complicated.
Hens play a vital role; if they are not good, the cocks will
not work properly. Moreover, the system takes too long to
implement. It is recommended that widowhood not be
considered for operational use at this time.
Disease
le' The following is a list of the symptoms, medication, and
methods of treatment for five of the commom pigeon diseases:
1. Canker (Trichomoniasis):
Symptoms� Failure of bird to swallow larger grains; swelling of the throat; cheesy
growths in the mouth area; loss of flesh and ambition, loss of appetite.
Navel in youngsters occasionally becomes infected and fills with the
cheesy deposits.
Medicine � Emtryl powder for flock treatment. Tricoxine tablets made by Fabry
for individual treatment.
Treatment� Powder: 1 Tbsp. per gal. of drinking water. Leave in loft for 5 days.
TCS 35974-77
149
Top cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
2. Coccidiosis:
Symptoms� Droopy, diarrhea (which may be bloody), lack of energy, loss of
weight, anemic appearance. Appetite diminished but not thirst. Loss
of ambition. Leg weakness, i.e., bird prefers to rest on heels rather
than stand.
Medicine � NFZ Soluble (powder).
Treatment� Add powder to drinking water according to directions on packet.
Keep treated water before birds for 2-3 weeks. Clean loft daily.
Change water at least once a day. Apply roost paint once every
month or so and install Vapona pest strips in loft.
3. Paratyphoid:
Symptoms� Old Birds: Loss of weight, decrease in appetite. Droopy, green, loose
droppings. Slight lopsidedness in flight. Swelling in wing and leg
joints.
Young Birds: Copious diarrhea; dizziness or evidence of brain
inflammation. Twisting the head sidewise. Disease worse in damp
weather. Caused by unsanitary conditions.
Medicine
� NFZ Soluble.
Treatment-
4. Roundworms
Symptoms�
Medicine �
Treatment�
Same as above. On old birds lance swellings and disinfect as of ten as
necessary.
(Ascaridia Columbae):
Droopy appearance. Loss of weight.
Piperazine Citrate. Also sold with an additive for hairworms
(Capillaria).
8 grams per gallon of drinking water of Piperazine Citrate over a
period of 60 hours. Scrupulous sanitation to prevent reinfection.
5. Pigeon Pox:
Symptoms� Wart-like lesions on unfeathered portion of body.
Medicine � Pox vaccine prior to infection.
Treatment� Vaccinate prior to disease. No cure.
TCS 35974-77
150
Top ret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
ecr et RUFF
APPENDIX D
A BRIEF HISTORY OF THE PIGEON
(U) Most of the following is taken from Chapter I of Reference
7, which is an encyclopedic treatise on every aspect of pigeons.
(U) The earliest recorded use of the pigeon (or dove) to obtain
information was by Noah, who sent out a dove knowing it would
return if it found no land. Instead, it did find land and returned
with an olive leaf, a demonstration of its love for home which
remains the most endearing quality of pigeons to this day. It is
speculated that King Solomon (about 1000 B.C.) used pigeons, but
the first documented evidence of their use in war begins with the
conquest of Gaul (over 2000 years ago) by Julius Caesar.
(U) During the War of Independence in Holland (1574), the
besieged people of Leyden were saved by messages of relief carried
by pigeons. Pigeons were also used by the Venetians during the
siege of Venice in 1849.
(U) It was during the siege of Paris (1870-71), during the
Franco-Prussian War, that the modern day Homer came into
international note. Balloons were released from Paris containing,
among other things, Parisian pigeons. These birds were retrieved
and taken to London, Tours, and other cities and subsequently
released with messages to the besieged Parisians. It was here that
one of the first uses of microphotography enabled the transport of as
many as 40,000 messages by a single homing pigeon. During the
four-month siege, 150,000 official and 1,000,000 private communica-
tions were carried into Paris by homing pigeons.
(U) In 1909, an international photographic exhibition was held
in Dresden, Germany. As invited delegates began their speeches,
pigeons with automatic miniature cameras harnessed to their bodies
made low-altitude photo passes over the exhibition hall. The exposed
film was quickly processed and converted into souvenir-postcard
enlargements for immediate sale to the delegates. This photo, and a
picture of the pigeon with camera, can be found on page 28, first
edition, of the -Manual of Remote Sensing,- Volume 1, 1975,
published by the American Society of Photogrammetry.
1
TCS 35974-77
151
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top ret RUFF
World War I
(U) The Germans developed military lofts as early as 1887 and
were quite prepared for the First World War, as were the French
and Belgians. It was the Belgians who first developed the modern-
day racing homing pigeon during the Industrial Revolution for the
purpose of carrying detailed messages of financial import between
England and Europe. Over 1,000,000 Belgian pigeons were taken by
the Germans during their occupation. To this day, many of the
better racing pigeons come from Belgium.
(U) It was not until 1916 that the first British birds were sent
to the front. British air force records show that 717 messages from
planes fallen in distress upon the seas were delivered by pigeons and
about 95 percent of several thousand pigeons came through with
messages. By the end of the war, British war Homers numbered
9,000 to 10,000 birds.
(U) When the United States entered the War, we had no
organized pigeon force. By 1916 birds were being trained to mobile
lofts. It was found that the birds soon came to recognize distinctive
markings on the roof of their lofts, which could be moved some
distance before their return. War Department records show, during
the Aisne-Marne offensive, mobile lofts enabled 72 birds to carry 78
important messages with no losses. In the St. Mikiel drive, 90
important messages were delivered by pigeons. Twenty-four of 202
birds were lost or killed, but every message was delivered since it
was sent in duplicate. In the Meuse Argonne offensive, 442 birds
delivered 403 messages safely from distances of 12 to 30 miles; not
a single message was lost. One bird, Cher Ami, was credited with
saving the -Lost Battalion,- and his body was mounted and placed
on exhibition in the Smithsonian. A second bird, -The Mocker,- was
awarded the D.S.C. as well as the French Croix de Guerre for
several outstanding feats of performance.
World War II
(U) The British were well prepared by the outbreak of World
War II. British breeders gave over 200,000 young birds to the
National Pigeon Service between 1938 and 1945. They were used by
the R.A.F. (standard equipment on all bomber and reconnaissance
planes) and the Army and Intelligence Services. Special Section of
the Army Pigeon Service (Secret Service) parachuted 16,554 birds
onto the continent. An outstanding example was the location of
German buzz-bomb sites. Pigeons were standard equipment for both
paratroopers and agents. Through the use of microphotography,
large quantities of plans and information could be delivered without
the severe risk of radio communication. The British furnished our
TCS 35974-77
152
ecret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
aet RUFF
U.S.A. Signal Corps, based in England, with 46,532 pigeons.
(U) Early in 1938, the U.S. Signal Corps had completed 20 lofts
for a new pigeon center at Fort Monmouth, New Jersey. Shortly
after Pearl Harbor, the War Department issued a call for champion
homing pigeons, one qualification being they must have flown 200
miles. The pigeon corps grew until, at its peak, it contained 54,000
pigeons, 3,000 enlisted men, and 150 officers. Major Otto Meyer, as
Commander of the Signal Pigeon Corps, supervised preparation of
the Army Technical Manual No. 11-410, -The Homing Pigeon," and
also Field Manual 11-80, -Pigeons for Combat Use."
(U) Pigeons were used extensively in the North African and
Italian campaigns. Here, pigeons were used by G-2 section (Intelli-
gence), and command posts who were so near the enemy that it was
impossible to string wire or use radio. They were also used by
armored patrols, night patrols, Ranger raids, etc. During the year
1944, the pigeons of the 209th Signal Pigeon Company serving with
the Fifth Army carried 10,286 messages. Of the 20,202 birds used
during this year, only 266 were lost.
(U) During the Luzon campaign, 2,594 messages were carried
by birds of the first Combat Platoon, 281 Signal Pigeon Company.
All messages were sent in duplicate, and not one was lost in spite of
mountainous terrain, rain, fog, hawks, and enemy shotguns.
Office of Strategic Services
(U) The 0.S.S. made outstanding use of pigeons in the Burma
campaign. One detachment, 0.S.S.S.U. 101, operated behind Japa-
nese lines in Burma and was commanded by Captain Morris Y.
Lederman. It was with this detachment Jungle Joe and Captain
Lederman achieved their renown. After only ten weeks in the
location, birds returned 225 and 250 miles when released by agents
who parachuted into the vicinity of Mandolay, Shwebo, and
Maymayo. The most outstanding flights were made by two five-
month old youngsters from a point near the Thailand border to the
loft at Bhamo. The distance was 325 miles.
(U) All agents, parachuted behind enemy lines, carried pigeons.
During January, 1945, nine groups were parachuted in and pigeons
either beat the radio or were the only means of contact for seven of
these groups. The distances flown were 175 miles, 225 miles, and
300 miles. The pigeons were held in jump containers from one to
three days. On another occasion, a pigeon was tossed from 150 miles
after 11 days on location and, although a resettled pigeon, it
returned in six and a half hours. A new shoulder message carrier
was developed, and pigeons flew 50 miles with a full roll of
111 TCS 35974-77
153
To. -cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
negatives. Out of hundreds of messages flown, only four were lost
and all were from distances greater than 150 miles. On several
occasions, pigeons were received from agents with urgent messages
for radio replacement parts.
Two-Way Pigeons
(U) These are pigeons which were trained to fly between two
lofts, eating at one and sleeping or drinking at the other. The U.S.
considered the method of training as Secret, though it is described
in the German Army Technical Pigeon Manual, published about
1925. The records of the 1308 Signal Pigeon Company shows they
flew two-way birds 55 miles.
Korea
(U) Pigeons were used by G-2 (Intelligence) of the Eighth
Army. During a four-month period, pigeons were used by seven
groups of agents parachuted from 75 to 200 miles north of enemy
lines. During this operation, not a single message was lost.
1
TCS 35974-77
154
To! .cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
cret RUFF
REFERENCES
2. "Naval Ocean Vehicles Material: USSR (U)," DST-180S-131-75;
1 November 1975; P. III-10.
3. Mitchner, M. and Walcott, C., "Homing of Single Pigeons Analysis of Tracks,"
J. Exp. Biol. (1967), 47, pp. 99-131.
4. Axelrod, Dr. H. R., "Pigeon Racing," Sterling Publishing Co., Inc., New York.
Sold by the Racing Pigeon Bulletin (see Ref. 12). One of several basic primers on the
care and keeping of homing pigeons.
5. Keeton, W. T., "The Mystery of Pigeon Homing," Scientific American,
December 1974, Vol. 231, No. 6, pp. 96-107.
6. Barker, Dr. W. E., "Pigeon Racing," 1958, The Racing Pigeon Publishing Co.,
Ltd., 19 Doughty Street, London, WC1. Sold by the Racing Pigeon Bulletin (see
Ref. 12).
7. Levi, Dr. W. M., "The Pigeon," Revised 1974, Levi Publishing Co., Inc.,
Sumter, S. C. Sold by the Racing Pigeon Bulletin (see Ref. 12). An encyclopedia
covering every aspect.
9. Whitney, Dr. L. F., "Keep Your Pigeons Flying," Paul S. Erikson, Inc., New
York. Sold by the Racing Pigeon Bulletin (see Ref. 12). An authoritative guide on the
subject of pigeon health, written by one of the foremost veterinarians in the country.
10. Schrag, Dr. L., "Healthy Pigeons," Verlag Ludwig Schober, D-8355,
Hengersberg, Obersteinhausen 66, West Germany, 4th ed., 1974. Recently translated
from the German and sold by the Racing Pigeon Bulletin (see Ref. 12). Color
photographs and a thorough approach to symptoms and treatment.
11. The American Racing Pigeon News (A monthly periodical), Thelma H.
Snyder, 2421 Old Arch Road, Norristown, Pa.
12. Racing Pigeon Bulletin (A weekly periodical), 94-G Compark Road, Center-
ville, Ohio.
TCS 35974-77
155
To. - -cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
cret
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 006527327
T ecret
c, NOFORN
k
1
Feasibility Research on a System to Provide High
Resolution Photography Over Denied Areas
(Supplement)
Handle via
TALENT-KEYHOLE Channels
ecret
RD-10-77
April 1978
TCS 35974-77
(Supplement)
Approved for Release: 2019/07/31 006527327
Copy NS 2
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 006527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
NOFORN
Feasibility Research on a System to Provide High
Resolution Photography Over Denied Areas
(Supplement)
A Research Study
By
Charles N. Adkins
Operations Technology Division
Office of Research and Development
APPROVED:
Chief, Oper
April 1978
s Technology Division, ORD Date
Handle via
TALENT-KEYHOLE Channels
TCS 35974-77 To ecret
E2 IMP
CL B�
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Top cret RUFF
NOFORN
This supplement contains high resolution prints of the Avian
Photography shown in the main report. The following table of
contents includes the page numbers of the main report where these
figures are printed at lower resolution.
TABLE OF CONTENTS
PHOTOGRAPHS
Page
2445 Color of Museum Park Facing page 1
10. Balloon Picture Using 3400 Film 24
14. Mobile Home Complex on Andrews Air Force Base ...... �� 31
15. Military Trucks on Andrews Air Force Base 32
16. Incinerator Plant on Andrews Air Force Base 33
17. Military Aircraft on Andrews Air Force Base 34
20. Corner with People Walking to Work 40
21. Alley Way 41
22. Roof Top with Air-Conditioner 42
23. Parking Lot 43
24. Oblique Shot of Museum Park 44
25. Navy Yard Main Gate 45
26. Old Naval Gun Factory 46
27. 2445 Color of Museum Park 47
28. Oblique Photograph Over the Navy Yard 48
A10-17 Examples of Vertical and Oblique Photographs
NIIRS Rated by NPIC Pis 106-113
A18-20 High Altitude Acquisitions 114-116
A21-22 Adjacent Frames Demonstrating the Effects of
Unpredictable Platform Motion (Frame-to-
Frame) 117-118
A23-27 Successive Frame Coverage of the Navy Yard
Target Area 119-123
A28 Example of Navy Yard Target Coverage by the
Model I (MCS-24) 7.5MM x 12.5MM Format
Camera, Not Mensurated ........ ............ .......... 124
A29 Ground Views of TALOS Missiles 125
A30 KH-8 Coverage of the Navy Yard Museum Dis-
play (Target) Area 126
A31 Color Photograph Taken by Tacana Over the
Washington Navy Yard, 7/8/77 127
A32(a-d) IDIMS Effect on Smeared Imagery 128
A33 Camera�Film Analysis Test Target 129
Handle via
TALENT-KEYHOLE Channels
TCS 35974-77 Top S
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
2445 Color of Museum Park
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE 10. Balloon Picture Using 3400 Film
Approved for Release: 2019/07/31 C06527327
A oved for Release: 2019/07/31 C06527327
-FIGURE -1/1-F Mobile Home Complex on Andrews Air Force Base
_ _ _
S ET
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
�
1-FIGURE 15. Military Trucks on Andrews Air Force Base SE
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
FIGURE 16. Incinerator Plant on Andrews Air Force Base'
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE 20. Corner with People Walking to Work
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
\ FIGURE 21. Alley Way I
ET
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
FIGURE 22. Roof Top with Air-Conditioner
Approved for Release: 2019/07/31 C06527327
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE 24. Oblique Shot of Museum Park
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
; FIGURE 25. Navy Yard Main Gate
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE 26. Old Naval pun Factory
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
(3,
�
\ A
,1
[FIGURE 27. 2445 Color of Museum Park
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
FIGURE 28. 'Oblique Photograph Over the Navy Yard '
_ _ _ _ _ - -
t tatio 1
"1141 (tA �
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
FIGURE A10. Washington Navy Yard-3/17/77
Rated NIIRS 9
REt
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
zvi
FIGURE All. Washington Navy Yard-3/17/77
Rated NIIRS 6
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
, FIGURE Al2. Washington Navy Yard-3/17/77
Rated NIIRS 7
RET
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A13. Washington Navy Yard-3/17/77
Rated NIIRS 7
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A14. Washington Navy Yard-3/17/77
Rated NIIRS 8
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
,FIGURE A15. Washington Navy Yard-3/17/77
Rated NIIRS 8
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
FIGURE A16. Washington Navy Yard-3/17/77 ,
Rated NIIRS 9
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
FIGURE A17. Washington Navy Yard-3/17/77
Rated NIIRS 9
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A18. Washington Navy Yard-7/7/77
Maximum Observed Flying Height
300 Ft. (91 M.)
NdFORN
Approved for Release: 2019/07/31 C06527327
roved for Release: 2019/07/31 C06527327
FIGURE A19. Washington Navy Yard-7/7/77
Maximum Observed Flying Height
300 Ft. (91 M.)
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
' FIGURE A20. Washington Navy Yard-7/7/77
Maximum Observed Flying Height
300 Ft. (91 M.)
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
LAP
^
�
FIGURE A21. Washington Navy Yard-7/7/77
Adjacent Frames (with Figure A22)
Showing the Effect of Platform
Dynamics on Image Quality
RET
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
FIGURE A22. Washington Navy Yard-7/7/77
Adjacent Frames (with Figure A21)
Showing the Effect of Platform
Dynamics on Image Quality
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
' FIGURE A23. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
ET
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
1
FIGURE A24. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FilGURE A25. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
ET
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A26. Successive Frame Coverage of the
Navy Yard Target Display Area
(Figure Series A23-A27)
NOFORNT
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A27. Successive Frame Coverage of the
( Navy Yard Target Display Area
(Figure Series A23-A27)
OFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A28. Navy Yard Display Area -
Not Mensurated
,T.
NOFORN
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 006527327
,FIGURE A29. Ground Views of Tabs Missile
at the Washington Navy Yard
Approved for Release: 2019/07/31 006527327
UNCLASSIFIED
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
FIGURE A31. The Washington Navy Yard-7/8/77/
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
32(a). Original Scene Micro-dSc
�
32(b). Rooting (Exp 0.7) Gaussian Filter!
32(d). Rooting (Exp 0.5) 5x5 Convolution, TTC
132 (c). Rooting (Exp 0.7) Gaussian Filter
Edge Enhancement _
FIGURE A32. !DIMS Effect on Smeared Imagery,
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 C06527327
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
FIGURE A33. Camera/Film Analysis Test Imagery
(From the Balloon Platform)
_
Approved for Release: 2019/07/31 C06527327
Approved for Release: 2019/07/31 006527327
Top ret
cret
Approved for Release: 2019/07/31 006527327