Closely-spaced-objects (CSO's) continues to be a critical issue for strategic infrared systems. Whenever a sensor's FOV encounters a high density of objects, tracking and discrimination performance can be critically affected by CSO's. The reason is that images (blur spots) on the sensor focal plane begin to merge as the distance between two objects is reduced. Some CSO's may be resolved by the signal processor, but more often the CSO is tagged without a resolution of the positions, amplitudes or number of objects from which it is comprised. We offer an approach, based on measurement of blurspot circumference which promises to advance the current performance limits for the CSO resolution function.
The nonstationarity of infrared interference backgrounds which prevents the implementation of the usual optimum linear filtering techniques makes clutter suppression signal processing for point target detection in infrared surveillance systems a challenging and difficult problem. Hence, more robust filtering schemes are sought which will perform well in structured backgrounds where the underlying probability distribution defining that structure is not well known or characterized. This paper investigates a promising candidate spatial filter for point-like feature detection in infrared systems. The technique, known as median subtraction filtering, is a robust, nonlinear, order statistic type filter which exhibits highpass filter characteristics without the usual ringing associated with linear highpass filters. A quantitative analysis of the statistical properties of the median subtraction filter is presented, including analytic expressions for the output distribution of the filter (thus analytic expressions for the probability of detection and probability of false alarm), its autocorrelation function and spectral density function. Performance results of a signal processing simulation comparing a median subtraction filter with an adaptive linear filter of the LMS type using actual infrared video as input are also included.
The detection of an unresolved point source in clutter is the stressing problem for Infrared Search and Track (IRST) systems. The availability of temporally spaced snapshots of the same clutter background results in a data source that is three dimensional: azimuth, elevation, and time. Here the clutter in different frames will be assumed to consist of two components: one that is constant from frame to frame and one that is statistically independent between frames. The component that is statistically independent is due to the combination of sensor noise, pixel position jitter, and changes in the background. The constant component represents the part of the clutter that is temporally unvarying; it can be reduced by frame-to-frame differencing. The component that is statistically independent between frames can be reduced by frame averaging. Each of these techniques is optimum (in the mean squared sense) only for an isolated set of conditions on the clutter and target behavior. The general optimum linear processor will be referred to as the three dimensional matched filter. Averaging and differencing are limiting cases of this filter: averaging when the target velocity is zero and the independent component dominates, differencing when the target velocity is large and the constant component dominates. The assumption of a constant clutter component and constant intensity target lead to a fillet with infinite temporal extent. The temporal extent of the three dimensional matched filter must be constrained in practice since only a finite number of frames can be stored. This constraint combined with the assumptions on clutter structure has lead to a computationally efficient implementation of the optimum linear processor based on a finite number of frames of data. This processor, along with methods for predicting and comparing its performance to other multiframe processing schemes, is presented in this paper.
Several simple, low-cost algorithms have been explored for use in the enhancement of imagery produced by uncooled focal plane arrays (UFPA). These algorithms address the main problems that UFPA-produced imagery typically demonstrate. In addition to enhancing UFPA-produced imagery, all these algorithms are simple and allow for inexpensive hardware implementation.
This paper will describe a simple technique that can be used to generalize the target classification algorithms employed by passive midcourse sensors for strategic defense. Most discrimination algorithm evaluations have assumed a fixed engagement geometry (target location/orientation, sensor location, sun and earth angles). Pattern classifiers are trained and tested in that geometry and therefore are not fully applicable in a full scale engagement. By training on the full range of potential engagements, the important class signature dependencies can be stored in an expanded mean vector and covariance matrix . Then through standard statistical techniques, the mean and covariance can be properly conditioned to the geometry applicable to a particular track file. This paper demonstrates that this approach is capable of adapting discrimination algorithms to a general scenario without significant loss in classification accuracy.
Early Automatic Target Recognizer (ATR) systems have been plagued by inconsistency in frame to frame recognition performance and high false alarm rate. One major source of this variation has been traced to instabilities in the Foward Looking Infrared (FLIR) image. The application of classical segmentation algorithms to the unstable FLIR imagery results in the extraction of a "wobbing" silhouette. This paper explores the attributes of first generation FLIRs that contribute to image instability and artifacts. The value of performing higher samples per dwell in a next generation FLIR sensor is explored along with the benefits of equal sampling in X & Y directions. Image instability and degradation result from a number of different sensor or environment factors that include aliasing due to undersampling, a-c coupling effects, 1/f noise, interface scan effects, mechanical scanner jitter, sensor platform motion, and atmospheric scintillation. To circumvent the sensor effects, the Army CECOM Center for Night Vision & Electro-Optics (C2NVE0) is currently developing an new generation of FLIR sensor under the SAIRS program. In parallel to the sensor development, two next generation ATR system testbeds are being developed under the Multi-Function Target Acquisition Processor (MTAP) program to support the evaluation of next generation ATR technology. These testsbeds include extensive instrumentation and are completely reprogrammable to facilitate the rehosting and evaluation of new ATR techniques. Following the MTAP thrust is a miniaturization program, ALgorithm Adaptive & Diminished DImensioN (ALADDIN), to shrink the size, weight & power of the current ATR processors. This paper presents an overview of the SAIRS, MTAP and ALADDIN programs within the context of the Army's overall plan for FLIR/ATR technology evolution.
The issue of infrared target imagery classifiability is investigated. Features that provide class discrimination information are automatically discovered and used to construct n-ary decision tree classifiers. Independent test data are classified and resulting performance reported. The importance of a priori information available in a realistic military scenario is assessed. Other factors affecting performance are also assessed. Probability of correct classification for a four-class problem ranges between 0.33 and 1.00 for various subsets of a standardized data base.
This paper presents a direct template matching approach dependent on an object's edge boundary profile. The method for matching an unknown signature to the prestored templates involves a minimum edge distance criterion. The classifier design details are preceeded by a derivation of the mathematical link between boundary and silhouette moments for binary objects.
Any electro-optical (EO) sensor operating in an environment in which there is a scarcity of information with which to discriminate one type of object from another needs to use all avenues available to it. For passive, nonimaging EO sensors, there are very few discriminants available; the spectral content of the radiation emitted by the object is one. The effective and efficient real-time use of this discriminant is, however, technologically difficult to implement. Recent studies have reviewed a wide range of possible techniques and have identified the relevant critical technologies involved. Thirty-two such concepts, listed here, have been developed over the years for a variety of applications both in the laboratory and in the field. Most of these are too cumbersome, too complex, or too slow to operate reliably and in real-time stressing environments. Two near-term techniques are multiple fixed filters and exchangeable edge filters. Two mid-term techniques are miniature dual-tunable Fabry-Perot (DTFP) devices and slit spectrometers. Two advanced techniques that use emerging technologies are electrically tunable detector materials and integrated-optics interferometers.
Bayesian probability theory is applied to the problem of the resolution of closely-spaced objects. The conditions assumed are: point sources, observed through a known smearing function (i.e., point-spread function). For this demonstration we use a Gaussian smearing function so that we can obtain analytic results; however, we present graphical results for both the Gaussian and the Airy smearing functions. The generalizations to arbitrary smearing functions may be found in other works by Bretthorst. The results obtained for one and two point sources indicate explicitly the dependence of resolution on signal-to-noise and on the smearing function.
A field portable IR Imager has been developed at Missile Systems Division of Rockwell International. The Imager is capable of extracting 128 X 128 and 256 X 256 images at 60 Hz rate. The system has been successfully demonstrated against a variety of ground and airborne tactical targets. Staring array used is MWIR HgCaTe hybrid operating in the 4.4-4.85 um region. The muliplexer is processed on silicon and is a CMOS switched FET device. The array is housed in a laboratory dewar and operated at 80K. A MRTD of .01K was measured with noise equivalent temperature difference (NEDT) of .03K. The array has been cycled between room temperature and 80K in excess of 40 times. The optics for the system consist of an F/2 8" lens giving it an IFOV resolution of .29 MRAD for 128 X 128 and .2 MRAD for 256 X 256 array. The Imager has the capability of varying its integration time between a minimum of .1626 msec. and maximum of 3.252 msec.
Several techniques are presented for modeling sky backgrounds in an infrared scene simulation. Each technique provides a part of the overall requirements for an accurate sky background model. A description of each technique is given and examples of each are presented. The main result of these investigations is that it is possible to combine both physically accurate models of atmospheric radiative transfer with computer algorithms for the generation of realistic cloud imagery.
The Modular Infra-Red Target Imaging Model (MIRTIM)is being developed at Rockwell MSD to assist in the design, development, and evaluation of IIR seekers and trackers. The goal of this effort is to produce a computer program that integrates target, weather, engagement, countermeasure, and seeker characteristics, including sensor and optics to determine anticipated seeker performance. The MIRTIM database will include 3-D ground and air target models with surface temperatures generated with the TACOM/KRC code PRISM, GTSIG from the Georgia Tech Research Institute, and Rockwell MSD's own AIRSIG Program. The weather effects database is obtained with the EOSAEL program library from the U.S. Army's Atmospheric Sciences Laboratory. Engagement, countermeasure, and seeker options are concerned primarily with tactical missile applications and focal plane array seekers. The Focal Plane Arrays are those developed by Rockwell International - Science Center, using their own Producible Alternative to Cadium Telluride Epitaxy (PAE) technology. These arrays have been shown to be an effective solution for tactical missile guidance. This modeling technology effort began in FY85.
Scene generation requirements are continually driven by increasingly complex threat and seeker/sensor hardware. Improvements in testing technology, which traditionally lag the development of new devices, have progressed significantly in recent years. The cost of extensive flight testing of surveillance and interceptor devices has given renewed interest to functional testing in a laboratory environment. Scene complexity has had to be greatly expanded to meet this need. Calibration curves are no longer sufficient to meet the growing needs of functional testing. This paper traces the historical development of LWIR sensor/seeker scene generation and projection technology. Electronic, all-digital and inband techniques are included.
McDonnell Douglas Corporation is involved in programs to improve survivability of electro-optic sensor systems that can be exposed to laser irradiation, from sources such as range finders and designators. The most vulnerable system components are the detectors. The detector damage level thus determines the system susceptibility. The system's detector susceptibility is tested initially with the detectors separate from the sensor system hardware. The McDonnell Douglas High Energy Laser Test Facility (HELTF) used for testing detectors is located in St.Louis Mo. It includes 23 lasers, computer controlled data acquisition systems, and thermal-vacuum test chambers. This facility has been used to characterize detector laser susceptibility for 13 years. Photo-conductive and photo-voltaic detectors have been tested from visible to long wave infrared wavelengths. The HELTF facility and detector susceptibility test techniques, including beam mapping and ensquared energy measurement techniques, are described in detail. Detector damage mechanisms are also discussed.
The use of lasers in range finders, target designators, and as battlefield weapons presents a problem for todays strategic and tactical electro-optical sensors. The sensitivity of these sensor systems makes them easy targets for jamming, spoofing, and damage by lasers. This directly impacts their survivability and the success of their mission. Hardening sensor systems against real and potential laser threats has been undertaken as a countermeasure. In order to harden sensor systems, their susceptibility must be assessed, hardening concepts incorporated, and finally, the hardening of the sensor must be verified. The McDonnell Douglas High Energy Laser Test Facility has been involved in testing vulnerability and survivability of sensor systems for the past thirteen years. The sensor vulnerability and survivability testing capabilities of the McDonnell Douglas High Energy Laser Test Facility are described. The lasers, instrumentation, and beam characterization capabilities as well as the threat simulation capabilities are discussed. Finally, test setups for several sensor systems tested in this facility are also discussed.
Testing of infrared systems such as FLIRs and missile seekers would be eased by the availability of an infrared scene generator with capability of displaying a complex, dynamic scene, the IR analogy of a motion picture projector. The two main categories of technologies being developed for this purpose are the spatial light modulator and the thermal emitter array. Honeywell is developing a thermal emitter array based on microstructure technology. This paper discusses performance parameters of scene generators such as resolution, radiometric accuracy, and time response. The number of pixels in an array may be an insufficient measure of resolution. Thermal emitters have advantages over light modulators when radiometric accuracy and high contrast are demanded. The unique physical and thermal parameters of the microstructure array technique help alleviate time response problems common to most thermal emitter arrays.
A technique was developed for fabrication of strong, fully dense, polycrystalline windows of CsI by hot-pressing in a steel die. We began with ultra-pure, ultra-dry CsI powder and conducted all processing operations within a glove box purged with dry N2. We were able to hot-press 5.08 cm CsI discs having transmittance of about 80% in the extra long wavelength (≥ 10 μm) infrared range--equivalent to that of single crystal CsI. The flexural strength of CsI bars fabricated by hot-pressing small-particle-size powder was 18 MPa, about 5 times higher than that of single crystal CsI.
The optical performance of physical vapor deposited beryllium mirrors replicated from tantalum-10 tungsten, molybdenum, and stainless steel master molds is being investigated. In this developmental replication process, a carbon release agent and then a layer of beryllium oxide are deposited on the master mold prior to beryllium deposition for crack-free mirror separation of the mirror from the master mold. The correlation of both optical figure and bidirectional reflectance distribution function (BRDF) between the beryllium mirrors and the master molds from which they were replicated is being experimentally determined. The reflectance of the mirrors replicated from the various master mold materials has been tested. Finally, thermal cycling of the replicated mirrors at cryogenic temperatures produces little change in BRDF values, but the optical figure changes indicate the need to remove the residual stress in the current replica beryllium mirrors. This technology development effort may lead to the fabrication of high-quality, low-scatter beryllium mirrors at a reasonable production rate.
We present experimental work performed on optical baffle materials to determine reflectivity changes in the materials following pulsed electron bombardment. Many optical systems require baffle materials capable of withstanding modest levels of low-energy particle pulses with little change in the surface characteristics of the material. Analysis of currently available baffle materials not only provides information concerning current optical system capabilities, but also enables understanding of baffle failure mechanisms which could help in the development of more durable baffle materials. Results show that most optical baffle materials are susceptible to pulsed electron bombardment such that the materials exhibit a change in surface reflectivity.
A high speed video data acquisition system was developed for aero-optical experiments conducted in a wind tunnel. The system acquired two dimensional video data at 370 frames per second. Each frame was an array of 128 by 128 pixels, each pixel digitized to six bits of intensity resolution. This produced a data rate of 8M bytes per second. To minimize the cost, off the shelf hardware was used, with the exception of a custom computer interface board. The design and operation of and future improvements to this system are discussed in this paper.
Measurement of optical coatings rely on spectrometers that are built primarily for measuring chemicals, where relative transmission or reflection is of prime concern. Spectral data for optical components on the other hand demand absolute photometric accuracy. The decision regarding what type of spectrophotometer system to use requires careful attention. Measuring transmittance through a wide range of refractive indices in materials can be challenging.
We describe the concept, limitations, design, and application of a broadband antireflection coating for germanium which does not use zinc sulfide or zinc selenide. These materials have some undesirable working properties in production and it can be shown that their index of refraction limits the spectral performance which can be achieved. A design using only germanium and thorium fluoride can meet a broad range of requirements when stress and environmental durability are kept under control.
The application of patterned electrically conductive (EC) coatings in infrared systems is discussed. The design approach, calculated and experimental performance for two different components are described. One component is a 9.5" diameter window of clear ZnS with an EC grid coating to provide deicing/defogging capabilities for a shipboard FLIR System. The heating uniformity and optical performance of the window were evaluated. The second patterned EC coating is an electromagnetic interference (EMI) shield on a window of ZnSe with a thin ZnS outer layer for improved mechanical properties with little loss of optical properties. The optical and shielding performance of this coated window for an aircraft thermal IR imaging system is reported.
Amorphous carbon films variously designated as i-C, α-C:H diamond-like or hard carbon films, have interesting optical and mechanical properties and are candidates for wear and corrosion protection in a wide range of applications. Because the depth of penetration by the indenter in normal hardness tests often approaches or exceeds the thickness of the diamond-like coating film, hardness data may reflect large substrate effects. This paper gives the results of ultra-low load microindentation hardness measurements which allow data to be obtained for indentation depths as small as 20 nm. The data also yield information on the elastic modulus of the coatings. Films were obtained from a number of sources and had been prepared by several techniques. The hardness values are in the range of 7 to 15 GPa and depend on the exact preparation parameters. The hardness data will be compared to those obtained by the Knoop technique.