Small target identification

Historic, Archive Document

Do not assume content reflects current
scientific knowledge, policies, or practices.

Reserve

t

aQC37 1

i qp! ™ted States

I cJOH- apartment of

griculture

I £7

Cl'

/Jj*- Forest Service i
Equipment
Development
Center ^

Project Record
8457 1202

5700— Aviation

San Dimas, CA

February 1984

Small Target** o^8
Identification

The Forest Service, U.S. Department of Agriculture has developed this information
for the guidance of its employees, its contractors, and its cooperating Federal and
State agencies, and is not responsible for the interpretation or use of this information
by anyone except its own employees. The use of trade, firm, or corporation names
in this publication is for the information and convenience of the reader and does not
constitute an endorsement by the U.S. Department of Agriculture of any product or
service to the exclusion of others that may be suitable.

Small Target
Identification

/

Robin T. Harrison -Aerospace Engineer
USD A— Forest Service
Equipment Development Center
San Dimas, CA 91773

Project Record
8457 1202

5700— Aviation

ED&T Project 2E-21P72-Small Target Identification

February 1984

CONTENTS

INTRODUCTION . 1

SIGHTS . 1

British Aerospace Corporation (BAC) Steadyscope . . . .1

Fujinon Stabiscope . 2

Kenlab Invisible Tripod . 2

THEORETICAL CONSIDERATIONS . 2

Optical Consideration . 2

Stabilization Systems . 4

TESTS . 5

Test Subjects . 5

Ground Tests . 5

Optics . 5

Evaluation of Sweep Stabilization Time . 5

Flight Tests . .6

Test Procedure . 6

RESULTS . .6

"Naive" Observers . .6

"Trained" Observers . .6

BAC Steadyscope . 6

Fuji 10X . .7

Other Systems . 7

CONCLUSIONS . 7

RECOMMENDATIONS . 7

ILLUSTRATIONS

Figure 1 . BAC Steadyscope test unit . . . 2

Figure 2. . Fujinon Stabiscope, 10X magnification model . .3
Figure 3. . Fujinon Stabiscope, 14X magnification model . .4
Figure 4. .Kenlab "invisible tripod” attached to binocular. .4

Figure 5 . Landolt C target . . .6

Figure 6 . Flight path . 6

ii

INTRODUCTION

The identification, observation, and counting of raptor
(hawks, owls, ospreys, eagles, etc.) nests, fledglings, and
eggs are accomplished by Forest Service wildlife biologists
as part of the eagle recovery effort, the protection of
endangered species, and widlife management in general.

The traditional way of obtaining such counts is by flying
over the nest trees at a low (less than 500 ft above ground
level [AGL] ) altitude in small fixed-wing aircraft. However,
this operation is in violation of Forest Service policy that
requires fixed-wing, single-engine aircraft to maintain an
altitude of at least 500 ft AGL for all operations. From an
altitude of 500 ft, the naked eye has been shown through
experience not to have the acuity necessary for performing
the required raptor observations.

Alternatives to direct observation, i.e., infrared detectors,
various varieties of cameras, time-lapse photography, etc.,
have also been considered. These alternatives are not
examined in this Project Record for several reasons.

All of the biologists contacted agree that the current
preference is for real-time visual observation. This is
because the parameters which the biologists are inspecting,
including counting of the young, looking for eggs and chicks,
etc., sometimes take as many as seven passes of the nest to
put the biologists into position to make the observations
needed.

Of course, the method which could reduce the number of
passes required would also reduce disturbance to the nest,
and this feature the biologists liked. The observer simply
needs real-time feedback to assure that accurate counts of
eggs and chicks, which are quite difficult to see against the
nest background, have been made.

Although an infrared detection scheme has been tried with
some success for large animals, no information exists on its
being applied to raptors. The opinion of wildlife biologists
is that even though such a system may have some application,
it would need an extensive period of development before it
could replace direct observation. Also, such systems are very
expensive, costing in the tens if not hundreds of thousands
of dollars, depending on the configuration and platforms
used.

One alternative to real-time direct observation would be
some sort of a stabilized television platform, equipped with a
television camera which has a remotely zoomable lens, and a
real-time monitor. Some observers have hypothesized that
a stabilized television camera with a remotely zoomable lens
could be passed once or twice over the nesting site. The

observer, sitting in the cabin of the aircraft, could see what
is being recorded on video tape via the monitor. A decision
could be made whether or not coverage is adequate. Once it
was determined that a satisfactory tape of the nesting site
was made, counts, measurements, etc., could be made from
the tape, played back on the ground where it could be
stopped for easy measurements, etc. A great deal of interest
was shown by the biologists in such a system. They were
particularly interested in the ability to "zoom" from "wide-
angle" to high magnification, as this would make the
counting and observation task much easier. The biologist
would spot the nest and instruct the pilot to maneuver the
aircraft to an advantageous position with the lens at "wide-
angle." The lens could then be zoomed in to obtain
necessary detail. If a system could provide adequate
stabilization for 12X magnification, without loss of
resolution, the observation task could be accomplished
from altitudes of roughly 1,000 ft, provided air speeds
could be held low enough to allow for target acquisition
in zooming while range was at its minimum. Such a system
will be tested early in FY 84.

Also, all of the biologists have shown a great interest in
stabilized real-time optical systems as the most acceptable,
currently available, alternative to low-altitude fixed-winged
flights.

The objective of the tests here described is to determine if
a stabilized, real-time optical system (hereinafter called
simply "sight") is available to do the raptor observation
job.

SIGHTS

Eight stabilized systems are, or have been, available. Of
these eight, only four appeared to have any promise at all
for our mission. These four are:

British Aerospace Corporation (BAC) Steadyscope

The Steadyscope is a monocular instrument, although its
body resembles conventional binoculars. It has two eye¬
pieces; one is blanked off. The example tested has a
magnification of 10X with a field-of-view of 6°.
Stabilization is accomplished by a gimbal-mounted
mirror, which is controlled by a battery-driven gyroscope.
The Steadyscope may be held in any attitude while in use.
Power is provided by a single "D" manganese alkaline 1.5-V
cell that provides 8 to 10 hr of running time. The Steady¬
scope weighs 4.4 lb, including the battery. Inspection of
the unit supports BAC's claim that the unit is a simple,
rugged, and dependable device. It is in use by the military
services of over 30 countries, including the United States.

1

The test unit (fig. 1 ) sells for approximately $4,900. BAC
has established a sales outlet in the United States at Dulles
International Airport near Washington, D.C.

Fujinon Stabiscope

The Stabiscope is available in two magnifications; both were
tested. They are similar in appearance to the Steadyscope—
see figures 2 (10X) and 3 (14X). The Stabiscopes also appear
to be rugged, well-built, precision units developed mainly for
the military market. There is a basic difference in the operat¬
ing principle, however. The Stabiscopes are true binoculars;
i.e., each has two complete optical paths, necessitating a
rather greater mass in the gyroscope gimbal system. This
results in a somewhat longer time for the gyroscopic mass
to stabilize after the binocular has received any angular input
than in the case of the Steadyscope. A small, external, re-
chargable battery pack powers each unit. It is not a standard
battery as is the BAC battery. Fujinon also has a U.S. outlet
(in Virginia). The price of the 10X is S3, 850; the 14X sells
for $4,250.

Kenlab Invisible Tripod

This unit is not a stabilized binocular, but is a gyro that is
attached externally to a binocular and stabilizes the entire
case. It uses an external rechargable power pack; the
stabilizing unit weighs 34 oz. The system is shown in
figure 4, attached to the binocular supplied by the
manufacturer. The KS 4 model tested, complete with
charger and power pack, sells for $2,1 1 7. Kenlab is located
in Connecticut.

THEORETICAL CONSIDERATIONS

In the evaluation of any stabilized optical system, two facets
must be considered: The resolving power of the optics and
the stabilizing properties of the gyro.

Optical Consideration

To be just detectable by a 20-20 eye with no astigmatism, a
target, with a 100 percent contrast against its background,
must subtend an angle of approximately 380 microradians
(prads). At a distance of 707 ft (the distance between the

Figure 1. BAC Steadyscope test unit.

2

Figure 2. Fujinon Stabiscope, 10X magnification model.

Figure 3. Fujinon Stabiscope, 14X magnification model.

3

Figure 4. Kenlab “invisible tripod" attached to binocular.

their nests) to make a precise calculation of the needed
resolving power. However, based on an analysis of a photo¬
graphy of a typical raptor in the nest, and on reports of
military users whose task has been to observe targets from
aircraft and other moving vehicles, it appears that a magnifi¬
cation of 10 would be optimum for our use. This estimate
is supported by the opinion of engineers from both BAC
and Fuji.

observer and the target in an airplane flying 500 ft above and
500 ft to the side of the target), the target, to subtend this
angle, would have to have a "smallest dimension" of approxi¬
mately 3.2 in. ("Smallest dimension" for the Landolt C
targets, used in our tests and below, is the gap in
the C, not the diameter of the targets.) Thus, with a 10X
sight, a target of roughly 1/3 of an inch should be detectable
from 707 ft.

In practice, many factors conspire to degrade this detect¬
ability. Imperfect optics, turbulence in the air, visibility-
reducing haze and dust, hand tremor, target movement,
contrast between target and background of less than
100 percent, target shape ambiguities, etc., all make the
ideal unachievable. In a moving vehicle situation, by far
the greatest limiting factor is the ability of the observer to
train the binoculars steadily on the target. For this reason,
increases in optical resolution (i.e., greater power) do not
improve the situation. In fact, they have just the opposite
effect as the higher the power, the greater effect of binocular
movement on resolution. Thus, one should select a system
that has the lowest optical magnification that will provide
for the necessary detection.

In our case, we simply do not have enough data about the
contrast between the intended ultimate targets (birds against

Stabilization Systems

Since any good optical quality 10-power sight will detect a
target as small as 1 /3-in at distances of interest in 1 00-percent
contrast backgrounds, and probably will allow for positive
identification of a 4- or 5-in minimum dimension fledgling,
even under contrast conditions of as low as 5 or 10 percent,
the limiting factor, in any stabilized binocular, will be the
stabilization system and not the optical system.

A basic consideration in a design of gyrostabilized systems
is that the lighter the stabilized mass, the more quickly it
"settles down," i.e., when the stabilized mass is subjected
to a rotational input, it takes a certain period of time to
come to a steady state. In a stabilized optical system, this
time period shows up as target instability. While this
instability persists, target detection and recognition
are impossible. The design of the BAC Steadyscope

4

features a very light stabilized mass. This is possible
because, among other considerations, a single optical
path is used. In the Fuji designs, parallel optical paths
are employed. Rather than a mirror, prisms are used as
the stabilizing element. These are necessarily heavier.

Also, because interpath alignment and rigidity are extremely
important, the stabilized mass, must of necessity, be heavier
than in a single-path system.

In the third system tested, the Kenlab, the stabilized mass
is much greater, including not just the binocular prisms,
but the entire binocular itself. Thus, the "settle-down
time" is much longer for the Kenlab than for either of
the other systems.

Brief evaluation of the "settle-down" time substantiates the
theoretical considerations, and indeed proved to be the most
important factor in the performance of the various sights.
This is dealt with below under "ground tests."

TESTS

A preliminary flight test was run from a Hughes 500 heli¬
copter. Three observers, all experienced aviation personnel,
used the BAC 10X and the Fujinon 14X and attempted to
observe ducks walking around on the ground from 500 ft
AGL. No quantitative test procedure was followed; however,
it was the unanimous opinion of the three testers that neither
system was optimum but that the BAC system might
perform our job. These tests were conducted at speeds
varying from a hover to 1 20 knots. In addition to observing
ducks, various features on the ground from 500 ft AGL were
also observed. While the opinion of the testers was uniform
that the Fujinon 14X optics were desirablejiall agreed that
the long stabilization time made successful observation
difficult, even from a hover. However, it was felt that a
test more closely simulating the actual raptor observation
task should be conducted before any sight should be
eliminated. Based on these observations, ground tests
and flight tests were planned and conducted.

Test Subjects

The same test subjects were used for ground and flight
tests. Six subjects were used, three "naive," and three
"trained." The "naive" observers were selected from a
pool of volunteers; all are employees of the San Dimas
Equipment Development Center (SDEDC). Their vision
was tested at a local optometrist. All six eyes had 20-20
vision uncorrected, five had no astigmatism; one had about
Vz diopter.

The three "trained" subjects were not all blessed with such
good vision. J, an SDEDC employee, does have 20-20 vision
with no astigmatism. R, Project Leader, is 20-30 with V/z
diopter in one eye and % in the other. G, who proved to
be the most successful observer, has worse vision yet;
roughly 20-40 uncorrected, but with only slight astigmatism.
As it turned out, the skill of the observers was much more
important than their visual acuity.

Test subjects R and J has considerable experience with
binoculars and other optical systems, although no previous
experience using stabilized binocular systems. Test subject
G was a trained Air Force navigator/bombardier who flew
in B-47's and B-52's. His observation technique was to look
for line discontinuity, not for the opening in the Landolt C.
He indicated that training was very important in the use of
any optical system and his performance in the test showed
this to be true.

Ground Tests

The objectives of the ground tests were to determine if the
optics of the sights functioned properly and to determine
what the "steady-down" time was for each of the sights.

Optics

We determined that the optical performance of all of the
sights was adequate; i.e., the optics did not limit the
detection job. At 500-ft sight distance, the position of a
0.45-in gap Landolt C target, 100 percent contrast, was
properly identified by all of the "naive" (20-20)
observers with near 100-percent consistency. Such a
target subtends an angle of approximately 75 prads so,
through a 10X system, the subtended angle would be 750
prads. This is approximately two times as large as the
theoretical detection limit of 380 pr ads. At the limit,
one would expect approximately 50 percent of the
"detections" to be accurate. Thus, we concluded
that the optical system would not limit the performance
of any of the sights.

Evaluation of Sweep Stabilization Time. This test was
done by sweeping the sights rapidly from left to right and
from right to left, and from down to up and up to down,
and stabilizing on a distant target. Results were obtained
from the BAC and both Fujinon units: The BAC stabilized
in less than Vz sec in all tests. Both Fujinon units were much
slower; stabilizing in roughly 3 to 5 sec. It should be noted
that Fujinon states that their binoculars should not be swept
more than 5 degrees-per-second. This is quite a slow rate of
sweep and it proved to be impossible to maintain while

5

attempting to acquire the target during the flight tests
described below. This rate was exceeded during these
sweep stabilizing time tests. A rotational rate of roughly
90 degrees-per-second was maintained during the tests.

This is approximately the rate of head rotation one
experiences while watching a tennis match.

The Kenlab unit proved to be impossible to test. The
concept of "settle-down time" had very little meaning
because the unit was so hard to train on the target. The
gyroscope is heavy enough to cause considerable procession;
i.e., when rotation is attempted in one direction, a gyro¬
scopic moment forces the sight in another. Several hours
of experience with the sight did not alleviate the problem.
Even after a distant target was acquired, small, slow panning
movements to inspect areas around the target center resulted
in unsatisfactory jitter and jump.

Flight Tests

The purpose of the flight test was to evaluate the stabiliza¬
tion system of the path of the sights by simulating as closely
as possible in a controlled manner the actual conditions
encountered during observation of raptors from fixed-wing
aircraft.

Test Procedure. The test was a controlled observation
of a Landolt C target by three "naive" observers and the
three "trained" observers, using the stabilized sights while
flying in a Cessna 182. The flight pattern was 500 ft above
and 500 ft to the side of the target. The C, shown in figure
5, was designed to have a contrast of 100 percent. Three
sizes of C targets were used. It should be noted that the
significant dimension in the target is the gap. The gap was
adjusted to one of eight positions so that the observers could
report the position to a technician on the ground. Radio
communication was maintained between the aircraft and
the ground crew. The test site is as shown in figure 6.

d

(in)

9

(in)

Large

8.0

1.6

Medium 4.0

0.8

Small

2.25

0.45

Figure 5. Landolt C target.

Initially, a ground speed of 120 mph was tried, however,
this was determined to be too fast for acceptable target
acquisition and so ground speeds of 75 to 80 mph were
used for the data gathering passes.

At all times the wind was less than 10 mph, turbulence
was never greater than occasional light. Visibility was
better than 60 mi, and there was no cloud cover.

RESULTS
"Naive" Observers

None of the "naive" observers could acquire the target
consistently. Even when they could acquire the target
through what: turned out to be the best of the systems,
the BAC, all three had zero percent correct answers for
the position of the target.

"Trained" Observers

BAC Steadyscope. The "trained" observers had no
trouble acquiring either the large or the medium Landolt C
while using the BAC system. The gap in the Landolt C on
the large target subtends an angle of 1,330 /trads in the 10X
scope, while the medium target gap subtends an angle of
670 jurads and the small target gap subtends an angle of
370 /trads.

None of the observers were able to correctly identify the
position of the small target, indeed none could state with
certainty that the small target was even seen. Since the
significant dimensions of the Landolt C are the width of
the limb (the black portion) and the width of the gap in

6

the C, this is not surprising (370 jurads is less than the
detection threshold of 380).

For the large target, observer G was 88 percent correct in
his positionings, observer R was 75 percent, and observer J
was 50 percent. This gives a clear indication of the stabiliza¬
tion performance of this sight, especially when compared
with the results for the other sights given below. For the
medium-sized target, only observers G and J made tests.
While observer G achieved a 58 percent (highly significant)
correct identification rate, observer J achieved only an
8 percent correct identification rate, which is insignificantly
greater than would expected by chance. This indicates the
overwhelming importance of training and technique,
especially remembering that observer J has 20-20, no
astigmatism vision, while observer G's vision is much
less than perfect. It is apparent from this last result that
a trained observer, even with less than perfect vision, can
detect a 0.8-in target with these binoculars under the
environmental conditions presented by this test.

It is the impression of ail three of the experienced observers
that even in the limited time that they flew with these
binoculars, their performance improved towards the end
of their test session. Time and budget unfortunately did
not permit further exploration of this point.

Fuji 10X. This sight came out a distant second best.
Neither observer G nor R were able to make any correct
identifications on even the large target. Observer J achieved
a 23 percent correct identification rate, which is greater than
would be expected by random luck. J was unable to success¬
fully acquire the target when the medium target was
su bstituted.

Other Systems. With the Fuji 14X, only observer G was
able to acquire the target and he returned a zero percent
correct identification rate. With the Kenlab system because
of the difficulties mentioned above, none of the observers
were able to even acquire the target area.

CONCLUSIONS

Of all of the stabilized sights tested, only the BAC Steady-
scope should be considered for further testing. This sight
will apparently meet the performance criteria necessary to
do our raptor observation job.

Training of the observers is absolutely necessary. Even
observers familiar with the job (i.e., observation of raptors)
need additional training, not just with the stabilized bino¬
culars, but with the stabilized binoculars used in the environ¬
ment in which the actual observation will be carried out.
Field experience as well as formalized training will probably
be required.

RECOMMENDATIONS

Since the BAC Steadyscope will apparently meet the
objectives of the project as set forth in the Project Plan,
further testing and development to lead to the implementa¬
tion of this sight should be carried out. This should include:

® Field tests using the BAC device to determine the
optimum techniques and field acceptability of these systems

@ The development of a training program and training
plan for the use of this sight in the raptor observation job

• Publication of an Equip Tips describing the BAC
Steadyscope and its proper use, and the training program
necessary to effectively use it.

7

02305

3172