Abstract: An apparatus for uniformly applying a compressive force against a filament to pneumatically drive the filament or fiber optic cable against a capstan without any concentrated amount of stress at any point on the filament is effectuated by pneumatically forcing the filament into an equatorial V-groove defined in the capstan. Pneumatic pressure is applied to a predefined segment of the capstan by a pneumatic shoe having an internal shoe pressure chamber. The pressurized gas or air is applied to the segment from the chamber within the shoe into the proximity of the equatorial V-groove on the equator of a disc shaped capstan. The V-groove is vented to atmosphere so that the cable is forced or blown into the V-groove. Side and end clearances between the rotating capstan and the shoe are sized to allow the viscosity of the pressurized gas to operate to retard the escape of the pressurized gas from the predefined segment of the capstan.

Abstract: An optical fiber canister (90) comprises a hollow housing (102) and an optical fiber pack (98) having a plurality of layers of optical fiber (20) supported on an inner surface of the housing (102) with a free end (96) of the optical fiber (20) positioned to pay out from an interior surface of the fiber pack (98). The optical fiber (20) of the optical fiber pack (98) has an amount of adhesive thereon ranging from zero to an amount sufficient to produce a peel force of less than about 2 grams. There is desirably a support layer (110) of a castable elastomeric material between the inner wall (108) of the hollow housing (102) and the outer surface of the optical fiber pack (98), and a release layer of a release material such as polytetrafluoroethylene between the inner surface of the support layer (110) and the outer surface of the optical fiber pack (98).

Abstract: A system (82) and method for locating features in an image. In the preferred embodiments, the present invention accepts as input a tilted view X-ray image of a PC board (10), as well as expected locations of solder joints (18) in the PC board (10). The present invention then determines the actual center locations (32) of these solder joints (18) by defining windows within the image and checking individual areas within the window, to see if they fall below a predetermined threshold. In addition, the system (82) determines if these pixels are connected to other pixels that have been previously determined to be part of the solder joint (18). Finally, the system (82) determines the center of the group of pixels determined to be part of the solder joint and displays the coordinate location of this center.

Abstract: An optical system (10) for producing dual fields of view simultaneously. The system (10) includes a first optical system (12) for producing a first field of view image and a second optical system (36) for producing a second field of view image where the angular displacement of the second field of view is different from that of the first field of view. A dichroic beamsplitter (22) is disposed in the present invention so as to reflect light from the first optical system (12). The dichroic beamsplitter (22) is also disposed so as to simultaneously transmit light from the second optical system (36). As a result, the reflected light (20) is primarily composed of light of a first wavelength band and the transmitted light is primarily composed of a second wavelength band. The light from the two different fields of view are then directed to a dual filter (32) which passes the first wavelength band in one portion and passes the second wavelength band in another portion.

Abstract: An optical fiber payout canister (22) comprises a bobbin (30) having a dispensing end and an optical fiber pack having a plurality of optical fiber layers (34, 36) wound upon the bobbin (30). The pack has a transition winding pattern between the optical fiber layers (34, 36) at the dispensing end thereof, and an overcoat adhesive layer overlies the transition winding pattern of the optical fiber pack. The adhesive layer desirably comprises from about 80 to about 72 parts by weight of a precatalyzed organofunctional siloxane polymer and from about 20 to about 28 parts by weight of a silicone elastoplastic resin. The adhesive layer desirably has a tensile strength of from about 25 to about 40 psi, and a modulus of elasticity of from about 1200 to about 1600 psi, over a temperature range of from about -50.degree. C. to about +80.degree. C.

Abstract: Two optical fiber segments are spliced in an end-to-end fashion by first axially aligning the optical fiber segments, and then fusing the optical fiber segments with a converging conical light beam convergently focused to an apex region along the optical fiber. The converging conical beam heats the optical fiber segments and the splice in a circumferentially uniform manner. The apex region at which the converging conical beam is focused can be moved progressively along the length of the optical fiber to effect the fusion, and also to directionally fire polish and stress relieve the optical fiber to minimize the presence of flaws in the optical fiber after fusion is complete. The converging conical light beam is achieved by creating a diverging conical beam using movable mirrors to deflect a collimated beam into a diverging conical beam. The diverging conical beam is reflected from a parabolic mirror to form the converging conical beam that is focused toward the optical fiber.

Abstract: A plurality of neural networks are coupled to an output neural network, or judge network, to form a clustered neural network. Each of the plurality of clustered networks comprises a supervised learning rule back-propagated neural network. Each of the clustered neural networks are trained to perform substantially the same mapping function before they are clustered. Following training, the clustered neural network computes its output by taking an "average" of the outputs of the individual neural networks that make up the cluster. The judge network combines the outputs of the plurality of individual neural networks to provide the output from the entire clustered network. In addition, the output of the judge network may be fed back to each of the individual neural networks and used as a training input thereto, in order to provide for continuous training. The use of the clustered network increases the speed of learning and results in better generalization.

Abstract: Method and apparatus for making a filament winding amenable to inside payout and having squared-off ends providing optimal volumetric efficiency. The filament crossovers are distributed so as to reduce crossover stacking in one or a few places which can interfere with winding and payout from the finished winding. A base wire layer is wrapped onto a mandrel over which a guide is nestingly wound with adjacent guide wire turns spaced apart. The filament first layer is wound nesting in the guide wire spaces, and subsequent filament layers accordingly formed. Each crossing filament turn has an advance in a crossing region substantially aligned with a crossover region in the underlying layer and which in one embodiment is one-half the winding pitch.

Abstract: A method and apparatus for detecting an object of interest against a cluttered background scene. In a first preferred embodiment the sensor tracking the scene is movable on a platform such that each frame of the video representation of the scene is aligned, i.e., appears at the same place in sensor coordinates. A current video frame of the scene is stored in a first frame storage device (14) and a previous video frame of the scene is stored in a second frame storage device (20). The frames are then subtracted by means of an invertor (24)and a frame adder (28) to remove most of the background clutter. The subtracted image is put through a first leakage reducing filter, preferably a minimum difference processor filter (32). The current video frame in the first frame storage device (14) is put through a second leakage-reducing filter, preferably minimum difference processor filter (36).

Abstract: A system for detecting phase and gain imbalance errors in a synchronous detector. The synchronous detector (10) is assumed to have a fist circuit (16) for providing a first signal representing a first sinusoidal term (e.g., a cosine term) and for providing a second signal representing a second sinusoidal term (e.g., a sine term) complementary to the first sinusoidal term, circuitry (12) for mixing an input signal with the first signal and circuitry (14) for mixing the input signal with the second signal. The system (16) for detecting phase and gain imbalance errors of the invention includes an amplitude compensation circuit (24) for detecting and correcting amplitude errors in the first and second signals and a phase compensation circuit (26) for detecting and correcting phase errors in the first and second signals. For amplitude compensation, the outputs of the amplitude and phase compensation circuits are input to first and second amplitude detector circuits (28) and (30).

Abstract: A frequency synthesizer (10) and method that achieves low phase noise and provides synthesized frequency signals in fine frequency step intervals. The synthesizer (10) comprises an L-band low phase noise frequency synthesizer with fine frequency step increments. The synthesizer (10) employs half-integer digital frequency dividing (25), VCO frequency offsetting, local oscillator harmonic mixing (34) and two phase locked loop circuits (12,13). The synthesizer (10) comprises a reference oscillator (11) for providing a reference frequency signal, and two phase locked loops (12,13). The first loop (12) is the fine loop and generates the frequency step size. The second loop (13) reduces the phase noise, reduces the frequency step size and translates to the desired freqeuncy. The output from a first VCO (21) is divided by a predetermined fixed number (31) to reduce the frequency step size and to reduce the phase noise.

Abstract: A "Venetian-blind" assembly of dichroic plates (36) is disposed in front of the entrance aperture (26a) of a Cassegrain-type telescope (26) which constitutes the optical focussing assembly in a tracking system for a guided missile (10) or the like. The plates (36) transmit electromagnetic radiation in a first optical wavelength band such as visible light, and reflect radiation in a second wavelength band such as infrared radiation. The plates (36) are inclined at progressively larger angles relative to the optical axis (22) of the telescope (26) such that the infrared radiation is reflected from a front surface (36a) of one plate (36) and subsequently from a rear surface (36a) of an adjacent plate (36) into the telescope (26) at a predetermined angle to the transmitted visible radiation.

Abstract: A technique for reducing the lateral force exerted upon a projectile launched into a flowing medium. The inventive technique includes the step of injecting a pressurized jet (J) into a medium (14), upstream of the launch point, flowing laterally relative to the anticipated path of a projectile (16). The projectile (16) is then propelled from a first location (18) into the medium (14) proximate the jet (J) injected therein.

Abstract: A circuit for reducing the commutation transient in a switch-driven (12) free-wheeling rectifier (20) applied to a load (10) via a continuously conducting energy conducting inductor (28), in which a further inductor (34) and diodes (32, 38) are interconnected with the switch-driven rectifier (20) to free-wheel recovered commutation energy from rectifier (20) when the switch is non-conducting. In other circuit variations temporary energy storage is aided by one or more capacitors (72) and additional diodes (52, 54, 80, 107, 122) during turn-on of rectifier (20), which energy recycled back into the circuit to minimize power dissipation and reduce circuit component heating.

Abstract: A lens assembly (30) focusses an optical image of a field of view (20) forward of a guided missile (12) onto a photosensitive target layer (50) of a television camera tube (52). The target layer (50) includes sublayers of zinc selenide (ZnSe) and cadmium telluride (CdTe)/zinc telluride (ZnTe), and has a spectral sensitivity range which extends for approximately 400 to 900 nanometers. A red filter (68) is provided to attenuate wavelengths shorter than 600 nanometers in the optical image to increase target contrast by reducing blue scattered light created by Rayleigh and aerosol scattering. The lack of response of the camera tube (52) to wavelengths longer than 900 nanometers further increases target contrast by reducing background radiation from soil and vegetation.

Abstract: A length of filament (32), wire or optical fiber, is wound into a pack (30) on a cylindrical drum (44) located within an open-ended housing (38). The housing is mounted within a missile (26) with the housing open end (38) facing aft. An undersized front wall (48) fits over the enclosure open end leaving a circular gap or space (50) through which the filament is dispensed. An outer end portion of the filament is reinforced (42) both mechanically and thermally to allow dispensing through outer regions of the missile rocket plume. The walls defining the circular gap (50) brake the filament dispensing to a predetermined maximum speed.

Abstract: A combination connector (20) includes a first half connector (22) with a fixed central body (40) and a slidable exterior housing (54) that is biased to a first position, and a second half connector (24) with a fixed exterior housing (70) and a slidable interior body (82) that is biased to a first position. The fixed central body (40) of the first half connector (22) has electrical contacts (42) on its periphery, and the fixed exterior housing (70) has electrical contacts on its interior wall. The slidable exterior housing (54) of the first half connector (22) protects the connection points before the connection is made, and is pushed back to expose the connection points as the two connectors are mated together.

Abstract: A birefringent prism (36) is disposed in front of the entrance aperture (26a) of a Cassegrain-type telescope (26) which constitutes the optical focussing assembly in a tracking system for a guided missile (10) or the like. The prism (36) refracts first radiation (O) having a first polarization in a first direction, and refracts second radiation (E) having a second polarization which is orthogonal to the first polarization in a second direction which is deviated from the first direction by a predetermined angle .DELTA..phi.. The telescope (26) focusses the first and second radiation (O,E) to form separate, laterally displaced first and second optical images (46,50) on first and second respective sections (34a,34b) of a focal plane photodetector array (34). Polarizing filters (56,58) which pass only the first and second polarizations therethrough are disposed in front of the respective sections (34a,34b) of the photodetector array (34) to eliminate optical crosstalk between the two images (46,50).

Abstract: The structural and gas-light bond between a radome and a fuselage in a missile cna be simultaneously fabricated in an improved methodology which reduces the numbers of fixtures required to be utilized, dramatically reduces the time for cure, simplifies the procedure, reduces the cost of fabrication and increases the yield. The fabrication process is comprised of the steps of applying an HT-424 film adhesive on the vertical flange of an etched surface of a fuselage. The film is spaced apart from the butt surface of the flange. A fillet of EA-934 paste adhesive is laid and formed into the corner of the flange between its vertical surface and the radially extending butt surface. The radome is then assembled onto the flange with all the paste adhesive which is extruded from the flange during assembly being immediately removed with a dampened cloth saturated with isopropyl alcohol.

Abstract: A high charge capacity readout cell in a hybrid focal plane detector array on a complementary metal oxide semiconductor integrated circuit chip. An input transistor that provides a buffer for the detectors of the array, couples to a source of bias voltage, which controls the operation of the transistor. An integrating capacitor uses a variable source of terminating voltage to increase the amount of charge it integrates. A read signal causes an output transistor to read the charge from the capacitor to a readout line and to initialize the capacitor. The termination voltage of the integrating capacitor is changed during the time that the detector current is integrated, thus increasing the change in total voltage across the capacitor. This allows a greater amount of charge to be integrated with the capacitor which improves the signal-to-noise ratio of the focal plane array.