 |
Description  |
|
|
FIELD OF INVENTION
This invention relates generally to radar systems and in particular to
air-to-surface missile guidance systems.
PRIOR ART
Most missile guidance systems use passive seekers located in the missile
nose for locking on a target. Passive seekers generally employ a camera
for sensing the radiation reflected from the ground. To-date laser,
electro-optical, and infrared cameras have been used in missile seekers.
The seeker camera is usually monitored on a scope in the aircraft by the
pilot before missile firing. When the pilot locates the target, he
maneuvers the plane so that the target is within the reticle of the seeker
camera. The pilot must then give a manual lock-on command followed by a
missile release command.
Upon missile launch, the seeker guidance system takes over. Either a
variable-size tracking gate or a correlation system for correlating the
present seeker camera image to a reference contrast pattern may be used to
control the missile servos and thus guide the missile to its target.
A major disadvantage of this type of system is that a high quality, high
resolution, wide field-of-view seeker camera must be used in order to
allow the pilot sufficient time (1) to recognize the target and (2) to
react so as to maneuver the aircraft so that the target is within the
seeker camera reticle before the plane passes over the target. This design
requirement leads to a very expensive seeker system for each missile.
Attempts have been made in the past to use seeker systems with just
adequate resolution and field-of-view properties. This trade-off of
resolution and field-of-view for a lower production cost leads to a high
aircraft attrition rate. This is because with a small field-of-view it is
much more difficult for the pilot to pick out a target from its
background. This problem is compounded when a low resolution system is
used. Since more time is needed to pick out the target and lock on it, in
any situation where the enemy has an up-to-date antiaircraft defense,
there will be a high aircraft attrition rate.
Thus there is a major need in present weapon guidance technology for a low
cost, air-to-surface, guidance system having sufficient resolution and
field-of-view to allow pilot to pick out a target against a cluttered
background at sufficient ranges to minimize aircraft attrition through
evasive maneuvers.
A second major problem with prior art systems is that only one missile
seeker camera may be monitored by the pilot at a time. Thus, where an
aircraft is carrying more than one missile and there are two targets in
close proximity to each other, the pilot is frequently forced to make
another pass over the second target. This procedure also leads to a high
aircraft attrition rate.
SUMMARY OF THE INVENTION
These guidance system problems are resolved in the present invention by
providing a small field-of-view, low resolution, forward-looking missile
seeker camera, a second forward-looking camera mounted in the aircraft
with a large field-of-view, a high resolution monitoring system for
displaying the second camera's large field-of-image to the pilot, and a
digital cross-correlator for locating the small field-of-view of the
missile camera within the aircraft camera's large field-of-view and
displaying this location on the monitor.
In this system, since the main monitoring camera is located in the aircraft
as opposed to the expendable missile, a much larger field-of-view, high
resolution system is economically feasible. Thus the pilot may begin to
monitor the target at much longer ranges, thus giving ample time for
target recognition, lock-on, and the initiation of evasive maneuvers.
Since the camera scene correlation is done electronically, no human
recognition factors need be taken into account. Thus a very small
field-of-view, low-resolution, seeker camera may be used. A desired cost
reduction is thereby accomplished by minimizing the complexity of the
expendable missile seeker at the expense of the aircraft mounted
equipment.
Since the aircraft camera's field-of-view encompasses a large area, a
number of small missile camera fields-of-view may be located on the main
aircraft monitor simultaneously, thus removing the second pass requirement
when targets are in close proximity.
OBJECTS OF THE INVENTION
An object of the present invention is to reduce the cost of missile-seeker
system.
A further object of the present invention is to increase the time the pilot
has for target recognition.
A still further object is to monitor the fields-of-view of a plurality of
missile seekers simultaneously.
A still further object is to increase the field-of-view and resolution of
the scene that a pilot actually monitors while decreasing the cost of the
over-all system.
A still further object is to decrease the aircraft attrition rate while
target boresighting by increasing the range at which boresighting may be
accomplished.
Other objects, advantages, and novel features of the present invention will
become apparent from the following detailed description of the invention
when considered in conjunction with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a schematic diagram of a typical field-of-view placement of the
mixxile and aircraft sensors.
FIG. 2 is a diagram showing an actual monitored image that the pilot might
see when the present invention is implemented.
FIG. 3 is a block diagram of boresighting system of the present invention.
FIG. 4 is a block diagram of a correlator that could be used in the present
invention.
FIG. 5 is a schematic illustration of an 88 .times. 64 element array that
may be used for the core memory.
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 illustrates a typical field-of-view placement in the basic system of
this invention. The aircraft 10 has a wide field-of-view, high resolution
camera 12 located in its nose. Its field-of-view 18 is shown encompassing
the target 22. A missile 14, slung beneath the wing of the aircraft 10,
has a seeker camera system 16 located in its nose. The seeker
field-of-view is shown as the dashed lines 20.
As can be seen from the drawing, the seeker system field-of-view 20 does
not encompass the target 22. Thus the pilot must maneuver his aircraft to
place the target within the seeker field-of-view.
FIG. 2 shows the actual monitored image that the pilot sees. The large box
24 is the high-resolution, air-craft cameras, sensed-energy image of the
scene. The dashed line box 28 represents the image that the missile seeker
camera 16 is sensing. The lines 26 represent the missile-seeker crosshairs
which must be placed over the target before the pilot may initiate the
standard lock-on procedure. It should be understood that only the
crosshairs of the seeker show up on the scope monitor 24.
A block diagram of the basic system is shown in FIG. 3. The missile sensor,
or seeker, for sensing the scene energy is the block 16. In the actual
embodiment developed, the sensor consists of infrared detectors (Ge: Hg or
(Hg-Cd)Te), although it is to be understood that any energy-sensing
detector over any frequency range could be used. Since human monitoring of
the seeker image is not essential in this system, the seeker is designed
to have only sufficient image quality and field-of-view to meet area
correlation requirements with the aircraft sensor 12. Thus only 8 infrared
detectors taking 16 samples per line, thus providing 128 resolution
elements, need be used. These detectors are arranged to cover a 1.degree.
by 1.degree. field-of-view. A 1 milli-radian resolution may be used. A
seeker with these parameters is the Aerojet Model T77 Seeker. This seeker
has 8 detector channels taking 16 samples per line thus providing an 8
.times. 16 image matrix. The seeker further has a rectilinear scan at 60
fields per second scan rate using a 2.1 interlace.
The aircraft scene sensor 12 is a forward-looking infrared sensor. Again it
is to be understood that any type of energy sensor over any frequency
range could be used. The actual sensor used is the Aerojet Model C19N
FLIR. The FLIR (Forward Looking Infrared sensor) has a dual field-of-view
capability; a wide field (20.degree. elevation by 25.degree. azimuth) for
acquisition and a narrow field (5.degree. elevation by 6.25.degree.
azimuth) for high resolution. Its resolution is 1 milli-radian for the
wide field and 0.25 milli-radians for the narrow field. The FLIR has 176
detector channels with 65 samples (resolution elements) taken from each
detector, thus providing a 176 .times. 64 image matrix. The detectors used
as Ge:Hg detectors. The correlator 30 is a digital cross-correlator. It
receives both the sensor 12 and the seeker 16 scene video outputs,
processes the video outputs for digitized storage, and compares the scenes
bit-by-bit for effecting an area correlation.
In order to allow the use of straightforward, digital, processing
techniques in the area correlator there are certain compatibility
requirements between the seeker 16 and the aircraft sensor 12. First, in
order to avoid the need for elaborate scan conversions in the correlator,
the seeker scan pattern must have approximately the same characteristics
as the aircraft sensor scan pattern. Second, it has been determined that
approximately 100 resolution elements are required to generate an adequate
cross-correlation function between two pictures while maintaining
adequately low side-lobe amplitudes. Thus the seeker 16 must have at least
100 resolution elements for each scene. Third, only elements of
approximately the same resolution may be correlated.
The display 32 provides a means for viewing the large field-of-view
aircraft sensor image. The section in this image that correlates with the
seeker image is distinguished by displaying a set of crosshairs at the
section center in the well known manner. This display 16 may consist of a
conventional cathode-ray-tube monitor with high resolution and high
dynamic range for optimization of the display/human eye interface.
In operation, the large field-of-view (20.degree. elevation by 25.degree.
azimuth) of the aircraft sensor 12 initially provides a sufficiently large
scene which can be examined at low resolution for examination for
potential targets. This scene is viewed by the aircraft crew via line 34
on display 32. When a potential target has been determined, a
high-resolution narrow field-of-view (5.degree. elevation by 6.25.degree.
azimuth) is used for examination on the display 36 and designation of the
target to be attacked.
At this time the correlator 30 is switched on, as is the missile seeker 16.
The correlator 30 digitally processes the scenes from the aircraft sensor
12 (5.degree. by 6.25.degree. field-of-view) and the seeker 16 (1.degree.
by 1.degree. field-of-view) and digitally compares these scenes
bit-by-bit. When correlation has been accomplished, a boresight error
signal (difference between the aircraft sensor crosshair coordinates and
the seeker crosshair coordinates) is sent via line 36 to the display 32.
This error signal is used to locate the seeker crosshairs within the
air-craft sensor scene being displayed on the display 32.
Thus the pilot of the aircraft using this display may maneuver the plane or
change the gimbal positions of the seeker detectors such that the target
is under the seeker crosshairs.
Alternately, a seeker-aircraft sensor slaving mode of operation could be
used. If seeker-to-aircraft sensor slaving is desired, the boresight error
may be applied via line 38 to the seeker 16 to activate the seeker's servo
system to reposition the seeker detector gimbals so that the seeker
field-of-view is centered on the aircraft-sensor crosshairs. If the
aircraft sensor is desired to be slaved to the seeker 16, the boresight
error is merely applied via line 40 to the aircraft sensor servos.
Activation of these servos causes the repositioning of the aircraft sensor
crosshairs to the seeker field-of-view crosshairs. Activation of gimbal
servo systems for repositioning is well-known in the art and thus will not
be discussed further.
A cross-correlation system that may be used in the present system is shown
in FIG. 4. The aircraft sensor inputs from the 176 detector channels are
shown as the numbered lines 50.
The initial 176 detectors (64 samples per detector) have a 1/4 mr
resolution. In order to make the sensor 12 resolution compatible with the
seeker 16 resolution (1 mr), the technique of averaging adjacent detector
channels is used. This produces an approximately 1 mr resolution in the
vertical direction. In FIG. 4 this averaging is accomplished by adding
adjacent channels through the resistors 53 in network 52. These averaged
channels are then filtered using the capacitor 55 in a low pass filter
configuration to match the sensor 12 resolution exactly to the seeker 16
resolution.
After this averaging, there are 88 channels forming an 88 .times. 64
matrix.
Generally some type type of multiplexing is required before this data may
be stored in a core memory. The particular type of multiplexing required
depends on the core memory used. In the actual embodiment developed, an 8
plane (i bit) core memory is used. Thus 8 inputs are possible at a time.
In order to handle the 88 channels, 8 multiplexers 54 with 11 inputs each
are provided. Each multiplexer 54 essentially comprises 11
field-effect-transistor switches. One switch in each multiplexer is biased
on such that 8 channel inputs are being applied to the memory at any one
time. After a short time interval, the next field effect transistor in the
set of parallel multiplexers is biased on. Thus these channels are sampled
8-at-a-time, until all 88 channels have been sampled. The multiplexer
switching is controlled by a timing control signal which is applied on
line 56.
After multiplexing, each channel signal is applied to a threshold detector
58 which converts the signals to binary 1's and 0's. In order to obtain a
proper correlation function, 50% of the picture elements must lie above
the threshold and 50% must be below the threshold of the detector. This
correlation requirement is met through the operation of a standard
adaptive threshold circuit 60. This circuit 60 merely determines the mean
level of the detector outputs using a set of comparing circuits and sets
the threshold in the detectors 60 accordingly via line 62.
These 88 channel inputs are then applied to the magnetic core memory 64 for
storage. As stated previously, this is a 64 .times. 64, 8-plane memory
with a read/write, full-cycle time of 2.5 microseconds.
When each channel has been scanned to obtain 64 samples and this 88 .times.
64 element matrix representing the FLIR sensor image has been
approximately sotred in memory 64, 8 .times. 16 blocks from this 88
.times. 64 matrix are systematically read-out and applied to a bit
comparator 80 under control of a memory address and control timing circuit
66. These 8 .times. 16 blocks are taken from the 88 .times. 64 image array
and compared in parallel to an 8 .times. 16 seeker image array in this
comparator 80.
The seeker-correlator interface will now be discussed. The 8 seeker input
signals 82 from the eight seeker detectors are applied to eight buffer
amplifiers 84. The buffer amplifier outputs are digitized to 1's and 0's
by eight parallel threshold detectors 86 (one for each channel). Again
there is an adaptive threshold circuit 88 identical to the circuit 60
controlling the detector thresholds by means of the line 90. Each of these
digitized seeker signals is applied as one input to a set of eight
parallel AND gates 92. The second input 94 into each of these AND gates 92
is a timing signal for strobing the digitized signals into eight 16-bit
shift registers 96.
The same timing signal 94 used to strobe the digitized signals is also used
to gate one of the eight shift registers in block 96 to the output line
98. This output line 98 contributes an input to the bit comparator 80.
A counter 100 counts the number of digital matches of these two parallel
inputs (one input representing the 8 .times. 16 block from the 88 .times.
64 matrix, one input representing the 8 .times. 16 element seeker image)
into the bit comparator 80. The correlation number (number of matches) for
each 8 .times. 16 block of the 88 .times. 64 array is applied to a count
comparator 104 to compare it to a number in the highest count storage
register 102. If this number in counter 100 is greater than the number
held in the count storage 102, it is read into register 102 as the new
highest count. The X and Y coordinates of this 8 .times. 16 block which
has the new highest count are read into an X-position update register 106
and a Y-position update register 108 respectively. This process is
discussed in detail at a later point.
The memory timing and address control circuit 66 is used to determine what
section of the 88 .times. 64 aircraft sensor map is to be compared with
the 8 .times. 16 seeker picture. Basically, the control circuit 66 reads
out of the memory 64 an 8 .times. 16 element array and stores this 8
.times. 16 array in an intermediate storage register 76 (eight 16-bit
shift registers). This 8 .times. 16 element array is then applied to a
bit comparator 80 in parallel with an 8 .times. 16 seeker element array.
The actual method of systematically taking 8 .times. 16 blocks from the 88
.times. 64 matrix is merely a question of formating and can be done in any
number of ways.
In this particular formating embodiment, 8-bit words must be processed
since an 8-plane core memory is used for the memory 64.
FIG. 5 illustrates the formating technique used on the 88 .times. 64 matrix
of the present embodiment. The memory 64 holds the 8-bit words in the
Y-direction as shown in the figure. Obviously, problems will arise when an
8 .times. 16 array extends over two 8-bit words. For example, if an 8
.times. 16 array has an X boundary extending from 1 to 16 and a Y boundary
extending from 2 to 9, two vertical 8-bit words must be used (the word
holding elements 1 to 8 and the word holding elements 9-16) in order to
obtain the 8 desired y elements.
In order to obtain the desired 8 elements, 2 consecutive vertical 8-bit
words are read from the memory 64 and applied to a 16-bit buffer shift
register 74. Eight of the flipflops of the 16-bit register 74 have outputs
to the intermediate storage register 76. When the 16 bits from the two
words have been shifted into the register 74, the timing control circuit
66 applies a set of pulses via line 68 to the register 74 to shift this
register until the appropriate 8-bits of the 16-bit shift register are
opposite the 8 output flipflops to the storage register 76. Thus in the
example, the register 74 would be shifted once so that the bits 2 through
9 are opposite the 8 output flipflops in register 74. These 8-bits would
then be applied to the intermediate storage 76 as one 8-bit row in the 8
.times. 16 array. Sixteen 8-bit rows would be read into the intermediate
storage 76 in this fashion. When the complete 8 .times. 16 array is read
into the storage 76, each element in the area is read out in parallel with
an element from the seeker 8 .times. 16 array and digitally compared in
the comparator 80.
The control circuit 66 consists of a set of two digital counters (a 64-bit
X-position counter 70 and an 88-bit Y-position counter 72) and appropriate
timing circuitry for operating the counters. The numbers held in these
counters represent the X, Y coordinate position of the bottom element of
the 8 bits that are to be taken from memory 64. For example if the array
with a position bounded by X = 1 to 16 and Y = 9 is desired, the first
number held in the counters will be X = 1, Y = 9. As stated previously,
when this number is read into the memory 64 the two 8-bit words X = 1, Y =
1 to 8 and X = 1, Y = 9 to 16 are read into the buffer register 74 and
appropriately shifted to obtain the 8-bits from Y = 2 to Y = 9.
The initial operation of this control system will now be described. The X
counter 70 is set to 1 and the Y counter 72 is set to 8. These coordinate
values are then applied to the X address and Y address in the memory 64
via lines 67 and 69 respectively. Thus these 8-bits (X = 1, Y = 1 to 8)
are applied to the storage 76 as the first 8-bit column in the 8 .times.
16 array in the before-mentioned manner. The X-position counter 70 is then
incremented by 1 and the next 8-bits are applied to the storage 76. Thus
continues until 16 columns of 8-bits are stored in the storage 76. This
8-x 16 array is then compared in parallel fashion to the seeker 8 .times.
16 array and the number of digital matches in the comparator 80 are
counted by the correlation counter 100. This correlation count is then
compared in a comparator 102 with the number held in the highest count
storage register 104. Initially this count in counter 104 is zero. The
comparator 102 thus determines that the number held in the correlation
counter 100 is greater and applies an enable signal to the gate 101 via
line 103. The count held in counter 100 is thus read into the
highest-count storage 102 as the new highest count.
It should be noted that the counters 70 and 72 always contain the address
of the lower-most left element in the 8 .times. 16 array during the actual
comparison process. Also it should be noted that the X-position counter 70
applies an input to the X-position update 106 via line 71 while the
Y-position counter 72 applies an input to the Y-position update 108 via
line 73. When the comparator 102 determines that there is a new highest
count, it applies an enable signal via line 110 to the update registers
106 and 108. Thus the numbers held in the counters 70 and 72 are read into
their respective update registers 106 and 108 as the coordinates of the
lower left corner element of the 8 .times. 16 matrix with the new highest
correlation count.
The intermediate storage 76 is set to retain the 8 .times. 16 array after
the comparison process by connecting each element in the 8 storage
registers of storage 76 for recirculation (feedback). Thus to compare the
next 8 .times. 16 array (X = 2 to 17, Y = 1 to 8), the first row in the 8
.times. 16 array (X = 1, Y = 1 to 8) is shifted out of the storage
register 76. Simultaneously the X-position counter 70 is incremented by 1
and this address (X = 16 + 1, Y = 8) is applied to the X and y memory
addresses. Thus the 8-bits (X = 17, Y = 1 to 8) are read into the storage
76 in the before-mentioned manner to form the new 8 .times. 16 array. This
new 8 .times. 16 array is compared accordingly. This process is repeated
until the X-counter reaches the count 64. At this point the X-position
counter 70 is set to 0 and the Y-position counter 72 is incremented by 1
to equal 9. This comparison process is continued until each 8 .times. 16
block in the 88 .times. 64 array is compared to the 8 .times. 16 seeker
array.
It is reiterated that this particular method of formating is merely one of
many that could have been used for memory storage and comparison.
When the 8 .times. 16 seeker array has been compared to every 8 .times. 16
block in the 88 .times. 64, aircraft, sensor image, the numbers held in
the X and Y position update registers 106 and 108 are transferred into
their respective position storage registers 112 and 114. These registers
112 and 114 serve as buffers for digital-to-analog converters 116 and 118
respectively. These D/A converters provide the analog boresight error that
may be applied to either the CRT display and/or the servo systems of
either the seeker or the aircraft sensor.
It should be noted that since the X-Y coordinates held in the registers 106
and 108 represent the lower-most left element in the 8 .times. 16 array,
the D/A converters 116 and 118 must be offset by a certain amount to give
the 8 .times. 16 array center (the actual seeker boresight). Thus the X
coordinate must be offset by eight to give X center coordinate while the Y
coordinate must be offset by four to give the Y center coordinate. The
actual boresight error (difference between the 88 .times. 64 array center
and the 8 .times. 16 array center being compared) is determined by
subtracting the 8 .times. 16 center coordinates from the 88 .times. 64
center coordinates (44, 32).
As stated previously, a major advantage in addition to low seeker cost is
that two or more seeker crosshairs may be displayed simultaneously on the
high-resolution, large field-of-view aircraft sensor. This may be simply
implemented merely by duplicating in the coorelator 30 the seeker
interface section (dashed line box 130) and the comparison section (dashed
line box 132) for each simultaneous seeker crosshair desired on the
display. An input from the intermediate storage 76 via line 81 could be
applied to each bit comparator 80 in the system. Each seeker crosshair
would then be summed into the video signal in the well-known manner. Thus
only one core memory 64 and 88 channel aircraft sensor interface is
required, thus providing a considerable cost savings.
It is to be understood, of course, that the aircraft sensor servos cannot
be slaved to the seeker boresight since there is now more than one seeker
boresight.
Obviously, many modifications and variations of the present invention are
possible in light of the above teachings. It is therefore to be understood
that, within the scope of the appended claims, the invention may be
practiced otherwise than as specifically described.
* * * * *
|
|
 |
|
 |
Description |
|
