Analog methods to deal with the nonuniformity correction problem for infrared focal plane arrays have been addressed. On-chip neural-network based techniques and simulation results of these schemes are presented. The approach is based on one in which the nonuniformity correction of individual pixels in accomplished by the method of steepest descent and its well known special case, the least-mean-square method (LMS). Alternative learning methods can be derived from the LMS method by considering only sign of the data or sign of the error. These nonlinear versions of the LMS methods have potential hardware implementation benefits such as power dissipation and simplicity. The simulation results of the nonlinear learning algorithm along with the LMS algorithm results are presented. Analog circuit elements implementing these learning algorithms are also presented. Our simulation results are very useful in that they model the nonidealities that are associated with specific hardware circuit implementation. These include the nonlinearities, scalar multiplicative error and offsets of both feedback multipliers and the integrators. The nonideal factors that cause instability and performance degradation in learning are multiplier and integrator offsets. These factors are shown to be much more important in the pixel offset correction than in the gain correction.

This paper reports on the performance of the Neuromorphic IRFPA, the first IRFPA designed and fabricated to conduct temporal and spatial processing on the focal plane. The Neuromorphic IRFPA's unique on-chip processing capability can perform retina-like functions such as lateral inhibition and contrast enhancement, spatial and temporal filtering, image compression and edge enhancement, and logarithmic response. Previously, all evaluations of the Neuromorphic IRFPA camera have been performed on the analog video output. In the work leading up to this paper, the Neuromorphic was integrated to a digital recorder to collect quantitative laboratory and field data. This paper describes the operation and characterization of specific on-chip processes such as spatial and temporal kernel size control. The use of Neuromorphic on-chip processing in future IRFPAs is analyzed as applied to improving SNR via adaptive nonuniformity, charge handling, and dynamic range problems.

Parameter variations cause unavoidable nonuniformities in infrared focal plane arrays and other integrated sensors. A one-time calibration procedure is normally used to counteract the effect of these variations between components. Unfortunately, many of these variations fluctuate with time--either with operating point (such as data-dependent variations) or with external conditions (such as temperature). Calibrating these sensors one-time only at the `factory' is not suitable--much more frequent calibration is required. We have developed an adaptive algorithm that continually calibrates an array of sensors that contains gain and offset variations. This paper extends the work of Ullman and Schechtman who developed an algorithm for gain adjustment. The adaptive nonlinear dynamical system can be mapped to analog VLSI or a discretized version may be efficiently implemented in digital hardware.

This paper discusses two applications of adaptive filters for image processing on parallel architectures. The first, based on the results of previously accomplished work, summarizes the analyses of various adaptive filters implemented for pixel-level image prediction. FIR filters, fixed and adaptive IIR filters, and various variable step size algorithms were compared with a focus on algorithm complexity against the ability to predict future pixel values. A gaussian smoothing operation with varying spatial and temporal constants were also applied for comparisons of random noise reductions. The second application is a suggestion to use memory-adaptive IIR filters for detecting and tracking motion within an image. Objects within an image are made of edges, or segments, with varying degrees of motion. An application has been previously published that describes FIR filters connecting pixels and using correlations to determine motion and direction. This implementation seems limited to detecting motion coinciding with FIR filter operation rate and the associated harmonics. Upgrading the FIR structures with adaptive IIR structures can eliminate these limitations. These and any other pixel-level adaptive filtering application require data memory for filter parameters and some basic computational capability. Tradeoffs have to be made between chip real estate and these desired features. System tradeoffs will also have to be made as to where it makes the most sense to do which level of processing. Although smart pixels may not be ready to implement adaptive filters, applications such as these should give the smart pixel designer some long range goals.

In this paper we further develop the theoretical foundations of the Wave Process and demonstrate several of its capabilities in three areas. This is a continuation of work that has been reported earlier. The equations describing the ideal Wave Process have been used to find parameter values that optimally match the response to a particular target velocity, for two different cost functions. The response of the ideal Wave Process has been determined for targets moving at other than the optimum velocity. Demonstrations show the Wave Process responding to maneuvering targets. Finally, the Wave Process equations have been extended to address targets moving against background. This ability to select the moving target and suppress the background is demonstrated for images derived from weather satellite photos.

The Neuromorphic IR FPA Sensor developed by Amber Engineering, Inc., Goleta, CA for the Wright Laboratory Armament Directorate's Advanced Guidance Division, performs a Difference of Gaussians filtering function, similar to what occurs in the outer plexiform layer of the primate retinal system. This function requires a computationally intensive (digital-wise) spatial-temporal data smoothing operation, which is executed on the focal plane, at the seeker frame rate, while the image data is still in the analog domain. Implementation of analog operation provides great flexibility, not only in terms of the speed and power dissipation advantages, but also with the interface of other processes to the analog system. The fact that the human visual system is essentially based upon analog techniques helps to emphasize the point; nature has invested millions of years in the development of sensors and processing `wetware' which are highly tuned to their environment. Our goal is to take further advantage of the lessons which nature can teach us, and advance the state of the art in imaging detection and tracking by taking the next step to develop neuromorphic/corticomorphic focal plane devices. This paper will discuss some of the concepts the authors have been investigation in formulating advanced `smart' FPAs for future guided missile seeker applications.

An on focal plane analog to digital conversion approach has been implemented for infrared sensor application. This development uses a patented oversampling methodology named MOSAD (Multiplexed OverSample Analog to Digital) in the design of simple circuits that can be placed at individual pixel sites. The construction of an analog to digital converter pixel is allowed with this technology. Most of the crosstalk and broadband noise associated with analog multiplexing and readout is avoided. Two demonstration designs were developed and built with Orbit, 1.2 micron CMOS Foresight process. For cost reasons, both designs were placed on the small die, 4.8 X 4.8 mm, and packaged in a 84 pin grid array carrier. These designs consist of a scanning array, 1 X 64 on 60 micron centers and two column portion of a 64 X 64 staring array on 60 micron centers. The detector buffer design will support HgCdTe high background applications. Support for the demonstration was received from Army, Night Vision Laboratory under their two color detector SBIR development program.

This paper discusses a signal conditioning and analog-to-digital (A/D) conversion principle suitable for infrared detector arrays, especially for applications using uncooled bolometer thermal detectors. An experimental 16 X 16 array has been designed including a columnwise A/D conversion. The A/D conversion method is of general interest whenever low cost, high digital resolution, and moderate speed is desired. High resolution and linearity is obtained without trimming. The technique is similar to a sigmadelta converter, but there is no need for complex decimation filters. Row-by-row readout operation of the bolometer array is supported by the columnwise A/D conversion. The 16-column preamplifier and A/D converter structure has been implemented in a standard 0.8 micrometers CMOS process, with 40 micrometers column pitch. The resolution is expected to be 16 bits with a conversion time of 78 microsecond(s) , and the power consumption is estimated to be about 0.5 m for a single column including preamplifier and A/D conversion. MOS transistor 1/f-noise is suppressed by electrical chopping at the preamplifier input. The on-chip columnwise A/D conversion has considerable potential for smart IR cameras with on-chip bit-slice processor architectures. A flexible single chip digital camera may be achieved with the implemented structure, since the digital column data of the A/D converter array and the rows of the detector array can be selected randomly. A theoretical analysis is made of how the SNR is affected by different levels of signal conditioning parallelism. This analysis can be used for qualitative comparison between different architectures.

Bias and signal conditioning techniques for uncooled resistance bolometer infrared detectors employ dc bias or various ac modulation and readout techniques. This paper describes a novel method of converting the absorbed incident radiation into digital signals by use of 0.8 micrometers CMOS oscillators in each pixel. Incoming radiation causes a resistance change in the bolometer which in turn results in a frequency shift. A concept of accurately readout out the digital signals from each pixel in the array is presented. Pixelwise A/D conversion and readout is demonstrated using a thin film semiconductor bolometer detector array. An NETD (Noise Equivalent Temperature Difference) of 0.1 K has been measured.

Uncooled pyroelectric arrays can be advantageously used for contactless measurements of one- and two-dimensional temperature fields. Satisfying values of noise equivalent power NEP, modulation transfer function MTF and long term stability of responsivity, respectively, are necessary for these applications. Linear pyroelectric arrays developed for those purposes are described. The pyroelectric chip based on lithium tantalate contains 128 sensitive elements (element size 90 X 100 micrometers ² with 100 micrometers pitch). A CMOS read-out circuit (low noise preamplifiers, S&H-stages, analog switching structures with digital components, output-amplifier) is specially designed. Pyroelectric chip, read-out circuit, and a PTAT temperature sensor chip, respectively, are mounted in a metal hermetic package with 8...12 micrometers germanium window. Measured NEP values reach 5 nW at a chopping frequency of 128 Hz. The modulation transfer function MTF (128 Hz, 3 lp/mm) measured is typically 60%. Devices for the measurement of temperature distributions based on linear arrays described contain the uncooled array, infrared optics, chopper, control electronics, analog to digital converter, and a comfortable digital processing unit for multi point pattern correction, accumulation, digital filtering and so on. The measuring range of such PYROLINE systems reaches from 0...80 degree(s)C, 50...300 degree(s)C, 200...700 degree(s)C (1500 degree(s)C), respectively.

A new analytical method to estimate temperature and emissivity for infrared measurements is described. There are four steps in the method. First, in calculating the output voltage, the dependence of temperature and emissivity of the object was evaluated. The result was the output voltage increased in proportion to the second power of the object temperature and the dependence of the emissivity was linear for the 250 K to 400 K temperature range. Second, in the fitting of these polynomial equations, the orders of six coefficients were also evaluated. Third, in measuring the output voltage of the standard imaging area, the unit transfer coefficient from digital unit (LSB) to voltage (V) was computed. Finally, an inversion problem for estimating temperature and emissivity of the object was proposed. We have developed a new kind of 3 approximately 5 micrometers band Schottky-barrier infrared CCD image sensor, which we call SCAT648, to verify the validity of the estimating method. The SCAT648 image sensor is composed of the different types of pixels. These pixels have different spectral responsivities and capabilities of measuring target temperature and target emissivity. Four standard temperature-controlled samples were imaged with the newly developed SCAT648 camera system. We estimate the error of the temperature and emissivity measurements to be a low +/- 0.5 K and +/- 5%, respectively.

Advanced infrared (IR) guided weapons, such as Theater High Altitude Area Defense and AGM 65, require a robust multispectral IR hardware-in-the-loop testing capability for high fidelity performance characterization and pre-flight and post-flight evaluation. The Kinetic Kill Vehicle Hardware-in-the-Loop Simulator facility, Munition Seeker and Evaluation Technology Branch (WL/MNGI), Eglin AFB, FL, provides the capability for the non-destructive performance testing of precision guided munition systems and subsystems. Particularly challenging is the development of scene projectors to stimulate IR imaging seekers with inband photons as an element of hardware-in-the-loop (closed-loop) guidance and control tests. IR scene projectors currently have limitations in radiometric power, resolution, response time, spectral band coverage, flicker control, and uniformity. Although today's projectors have overcome some of these limitations, the advances have been achieved through technology trades such as flicker control and spectral band coverage in laser projectors and radiometric power, flicker control and response time in the liquid crystal light valve projectors. WL/MNGI is developing a spectrally tailorable IR scene projector under the WISP program that overcomes the need for the limiting trades of the past, providing the most robust and flexible design to date.

The Scene Generation Test Capability (SGTC) program at AEDC has completed development of the Direct Write Scene Generation (DWSG) test facilities which provide a dynamic mission simulation capability for Focal Plane Arrays (FPAs) and their associated signal processing electronics. The first phase of the program was completed in September 1991 and supplied a near-term test capability (designated for Transportable Direct Write Scene Generator, or TDWSG) to meet the test requirements of future early warning sensor systems. Over the last two years the TDWSG has been involved in test activities to validate the DWSG technique for meeting system mission simulation requirements. The DWSG approach is based on the ability to accurately control the position and amplitude of multiple laser beams through the application of radio frequencies to a set of acousto-optic deflectors. One of the primary concerns related to using the TDWSG for mission simulation is the system noise associated with the test facility. A system noise study was conducted using a low noise LWIR Si:As FPA and an optical power meter. Radiometric signal and noise measurements were acquired and used as a baseline for comparison with the TDWSG data to quantify the noise contributions of individual TDWSG subsystems. This paper presents an overview of the DWSG concept, results of the system noise study, and results of system precision measurements.

The Arnold Engineering Development Center (AEDC) Scene Generation Test Capability (SGTC) program has completed the development of a laser based Direct Write Scene Generation (DWSG) facility that provides dynamic mission simulation testing for infrared (IR) Focal Plane Array (FPAs) and their associated signal processing electronics. The AEDC DWSG Focal Plane Array Test Capability includes lasers operating at 0.514, 1.06, 5.4, or 10.6 micrometers , and Acousto-Optic Deflectors (AODs) which modulate the laser beam position and amplitude. Complex Radio Frequency (RF) electronics control each AOD by providing multi-frequency inputs. These inputs produce a highly accurate and independent multi-beam deflection, or `rake', that is swept across the FPA sensor under test. Each RF amplitude input to an AOD translates into an accurate and independent beam intensity in the rake. Issues such as scene fidelity, sensor frame rates, scenario length, and real-time laser beam position adjustments require RF control electronics that employ the use of advanced analog and digital signal processing techniques and designs. By implementing flexible system architectures in the electronics, the overall capability of the DWSG to adapt to emerging test requirements is greatly enhanced. Presented in this paper is an overview of the signal processing methodology and designs required to handle the DWSG requirement. Further, electronic design techniques that enabled the system to be implemented within program cost constraints will also be presented. These electronic designs include a broad range of disciplines including digital signal processing hardware and software, programmable logic implementations, and advanced techniques for high fidelity RF synthesis, switching, and amplitude control. Techniques for validating electronic performance will also be presented along with data acquired using those techniques.

The Arnold Engineering Development Center (AEDC) has developed an extensive focal plane array (FPA) test capability that can be used to measure performance at the detector array or hybrid array level. The test capability has been developed around an integrated approach to sensor testing. A suite of test facilities has been developed to allow both radiometric calibration to be performed and mission simulation issues to be evaluated. The test facilities can be interfaced either to the FPA or to the array with its accompanying signal processing electronics. More than 100 FPAs of various sizes and types have been characterized to varying degrees in the AEDC FPA test facilities. Arrays designed for both strategic and tactical applications have been characterized. Numerous reports and data packages have been produced to document these tests. Extensive analysis of data acquired has been performed for a large number of arrays.

The high performance requirements for the Theater High Altitude Area Defense (THAAD) Seeker required the build and verification of a state-of-the-art infrared seeker test and evaluation facility. The test and evaluation facility is completely enclosed in a class 10,000 clean room and is divided into four major areas. These areas are the build and assembly area, goniometric test area, boresight test area, and analysis area. The build and assembly area is where parts are inspected, cleaned, kitted and finally assembled. After assembly is complete, the seeker is moved to the goniometric and boresight test areas for calibration and test. The goniometric/radiometric test area is where seeker gain and offset, IFOV, FOV, FOR, PSF's, dynamic range and uniformity tests are performed. The boresight test area is where the seeker boresight and servo rate tests are conducted. The seeker operation and performance is controlled and monitored via the Seeker Test Set (STS). The STS provides seeker power, controls all seeker functions, collects simultaneous servo and image data and controls table movements and blackbody target temperatures. For storage and further analysis of data, the STS has been networked via an ethernet connection to the data analysis area. The analysis area contains an off-line data processing and reduction lab consisting of networked high performance PC's. This paper discusses the test facility created for the THAAD IR seeker including requirements, layout and unique functionality.

The Arnold Engineering Development Center (AEDC) supports ground environmental testing with a variety of versatile ground test facilities. The AEDC mission is to test and evaluate aircraft, missile and space systems and their subsystems. This paper will focus on the facilities available and under construction at AEDC to meet the space ground test needs of government and commercial customers. The space chambers and their associated test methodologies are divided into the following areas: sensor calibration and mission simulation, nuclear weapons effects, thermal vacuum/solar simulation, contamination, and component checkout testing.

The infrared measurement facility at the NRaD (formerly NOSC) division of NCCOSC has been involved in infrared radiometric testing for over forty years. This versatile facility provides complete radiometric testing of single element detectors, detector arrays, and hybrid focal plane arrays. Measurement capabilities include spectral response, flood illumination, spot illumination, and ionizing radiation testing, in low to medium background photon flux environments.

IR imaging systems require correction of detector to detector gain and offset differences to obtain high quality imagery. Ideally this correction would be performed once at the factory and would remain stable indefinitely. For most imagers, this ideal is not obtainable and some level of correction must be performed. The stability of the gain and offset characteristics of the FPA are therefore important to understand as it affects how often this correction must be updated.

With increased requirements for better performance being placed on thermal imaging systems, new characterization figures of merit are being developed to assess infrared focal plane array (IRFPA) attributes. Post correction uniformity (PCU) is a parameter that determines how successfully a thermal imaging system can eliminate spatial noise from scanning and staring focal plane arrays. Requirements on PCU, particularly for the more sensitive IRFPAs and applications, are quite rigorous. Test issues of l/f noise, drift, and repeatability become critical and require a rethinking of accepted methods. As infrared sensors have become more sensitive, the need to characterize these focal plane arrays under more controlled and realistic test conditions has emerged. The U.S. Army Night Vision and Electronic SEnsors Directorate (NVESD) has attempted to address these issues by developing a unique capability to measure the PCU of IR focal plane arrays using software algorithms and a specialized mechanical modulator. The modulator is a two foot diameter, two toothed (one reflective and one emissive) blade, which is used to facilitate the real-time collection of test, gain, and offset flux levels. This paper addresses (1) the significance of PCU from a system perspective, (2) discuss the limitations of various PCU measurement techniques, (3) present the NVESD approach for measuring PCU, and (4) report PCU data collected using these techniques.

Intelligent decisions on the selection of infrared Focal Plane Arrays for sensor systems can only be made if testing of the individual detectors/arrays is carried out in a systematic and accurate manner. Devices are often tested several months apart, in different locations and under varying radiation conditions. This paper discusses various experimental methods that can be used to insure that consistent, accurate results are obtained. Radiometric parameters must be carefully controlled. Blackbodies make excellent reproducible infrared sources when used correctly, as they can be calibrated with simple thermocouples. Selection of blackbody temperature for various wavelengths of interest is important since measurements in a poorly chosen spectral region can translate minor blackbody temperature fluctuations into large errors in detector photon flux. Design of Dewar geometry is extremely important for low background measurements, as large errors can occur due to unwanted reflections or light leaks. The physical dimension of any limiting apertures must be measured precisely. Filter characterization must be accomplished for each filter or window used at the temperature of interest. With care, systematic and accurate measurements, and hence valid comparisons, can be made between detectors measured at different sites and times.

The paper describes the establishment of an infrared spectral responsivity scale at NPL with an uncertainty of 1.6% and the development of facilities to characterize the optical properties of infrared detector, arrays and cameras. The paper will illustrate the use of the calibration facilities with examples of uniformity, linearity and spectral responsivity measurements on a range of detectors.

We describe a low-cost personal computer-based generator designed to produce timing signals needed for the operation of visible or infrared focal plane devices and other electronic systems requiring a large number of complex parallel timing control signals with precise relative phase relationships. The implementation is a high-speed static memory digital board which can be programmed by copying the timing waveforms emanating from a commercial generator (e.g. Pulse Instruments PI-5800A, Interface Technology RS-690, etc.). This simple approach retains most of the desirable features associated with using a commercial timing generator such as flexibility and the use of user-friendly software for timing synthesis. The approach permits the implementation of timing generation to multiple focal plane devices, each operated by a unique local copying and replicating timing board, from timing data synthesized on a central commercial timing generator. As evidenced in our laboratory, this approach has resulted in a lower cost for operating multiple focal plane devices with little loss of flexibility and programmability.

A low cost, PC based Infrared Focal Plane Array (FPA) test station has been designed and implemented utilizing state of the art analog to digital converters, a digital signal processor, and Windows based software. A modular design allows for ease of upgrades. New A/D's, additional memory, or state of the art digital signal processors can be implemented without making major changes to system hardware or software. The software accommodates all of the various scanning and staring arrays made at Texas Instruments. The modular design of the software minimizes major rewrites when new FPA configurations are encountered. The overall laboratory test dewar (FPAs must be cooled to cryogenic temperatures for operation and testing) based testing philosophy has also been modified. In order to facilitate a low noise design a lab dewar that simulates a tactical dewar (the final IR system integrated cooling package) with side boards has been used. These `side boards' eliminate any unnecessary cabling to and from external hardware, thereby reducing pick-up noise. The test station's noise floor routinely approaches the theoretical RMS noise limit for the various A/D's being used (16 bit 2 MHz and 14 bit Mhz).

Rockwell has combined two independent, lower-cost testing approaches proven out on the U.S. Air Force's Manufacturing Technology for HgCdTe Focal Plane Arrays (MANTECH) program and the Ballistic Missile Defense Office's (BMDO, formerly Strategic Defense Initiative Office (SDIO)) Hybrids With Advanced Yield for Surveillance (HYWAYS) program. These two testing approaches are, respectively, cryoprobing focal plane arrays and using multiple device test dewars. These approaches will be combined for significantly lower testing costs for defense-related programs such as the Air Force's AGM-130 program. The revised methodology utilizes the low-cost approach of cryoprobing batches of up to nine focal plane arrays, mounted on test carriers, to screen out the predominant failure causes, followed up with final acceptance testing of the arrays passing the cryoprobe test in multiple device dewars. This approach is projected to reduce the testing costs by a factor of 1.6 compared with the current testing approach of single device dewar testing. Planned improvements of increasing the degree of automation and improving the cryoprobe station to measure noise are expected to further reduce testing costs by an additional factor of six.

Second-generation technology production and development testing must be performed by equipment and processes that are capable of handling the tasks in an economically efficient manner. As such, data acquisition and reduction times, configuration change complexity, and test set recurring costs must be kept at a minimum to meet the needs of the second-generation IR factory. The maximum test throughput must be achieved, while meeting all technical requirements, using a minimum of program or capital assets. SBRC's method to accomplish this includes the design of the next generation of infrared test station, with a defined interface architecture, that allows great flexibility in the use of optical tables, warm and cryoprobers, and other test equipment. The paper will present a comparison of relative cost and capability between this most recent generation of test stations and the past generations. Benchmarks of key data acquisition and reduction speeds will be discussed. Also, benchmarks of configuration change time and performance may be included. The design of the interface architecture that allows flexible use of all supplemental test equipment (such as optical tables) is addressed. A general comparison of pre- and post-test equipment changes, as they relate to test throughput on a macro level, is included. There will also be a discussion of the increased capabilities of IR development and production test at this facility.

We describe in this paper a flow process that details design, fabrication, and test methodologies for the production of linear longwave infrared (LWIR) HgCdTe (MCT) detector arrays. The modular manufacturing approach emphasizes testability in the component design and zero-loss procedure in the FPA assembly that achieved reliable, low-cost production of high performance scanning LWIR focal plane arrays. In-depth theory, practice, and automation of infrared detector array evaluation are also discussed.

This paper will describe the USAF Arnold Engineering Development Center (AEDC) technology efforts that provide signal processing and data system support for infrared (IR) Focal Plane Array (FPA) testing. The requirements for AEDC space sensor testing range from component-level FPA characterization to advanced mission simulation. The technology efforts underway address these requirements by developing hardware and software that meet AEDC's generic needs for FPA testing. Component-level FPA characterization places unique requirements on system fidelity and bandwidth performance. Diversity in sensor types being tested and levels of sensor integration creates the need for versatility in data handling and sensor interfaces. Mission simulation requirements emphasize the need for extended data storage, system throughput, and data display capabilities. A signal processing system will be presented which addresses AEDC's requirements for component-level sensor operation, data acquisition, and flexible interface architectures that can be modified quickly to accommodate different sensor interfaces and data formats. The system will also address the need for high- speed storage of very large data arrays during mission simulation testing. Techniques used to verify and validate system operation will also be presented.

Low-leakage silicon p-i-n diodes have been investigated for gamma rate dosimetry. Radiation- induced current response was measured versus gamma flux rate in the range of 10⁷ to 10⁹ gamma-photons/cm²(DOT)s. Energy deposition dose rates inferred from radiation-induced current agree well with expected results based on gamma energy-absorption coefficients. Degradation of diode leakage current due to total dose was also tested.

The fluence of pulsed lasers of wavelength 4 and 10.6 microns necessary to induce one and two orders of magnitude temporary degradation in the R₀A values of Hg_0.7Cd_0.3Te p/n infrared detectors at 100 K, and Hg_0.78Cd_0.22Te p/n infrared detectors at 40 K have been calculated. A nonparabolic energy-momentum relationship and temperature dependent energy gap of HgCdTe were used in this calculation. The R₀A values used in this calculation were obtained by simultaneously including generation-recombination, diffusion and tunneling mechanisms.

The Focal Plane Characterization Chamber (FPCC) at the Arnold Engineering Development Center is configured to provide highly accurate radiometric characterization of focal plane arrays (FPAs). The chamber offers several inherent advantages that render it a unique and highly versatile facility. In addition to the excellent radiometric calibration capability provided by the FPCC, the chamber contains systems to allow several special FPA performance issues to be evaluated. Four of the principal capabilities that have been implemented provide the capability to evaluate performance issues such as crosstalk, FPA response blooming, radiometric flash recovery, and response in various spectral bands. These specialized test capabilities are described and discussed herein.

A system for measuring the average relative spectral response of long wavelength infrared focal plane arrays (IRFPA) has been developed. Typically, the spectral response of sister detectors, fabricated on the same wafer as the FPA detectors are characterized in lieu of characterizing the actual IRFPA and it is generally assumed that the IRFPA detectors are equivalent to the sister detectors. On occasion, however, this assumption has proven to be incorrect. The spectral capability described here was implemented by interfacing an IRFPA test station to a long wavelength monochrometer, allowing the spectral response of the detectors in the IRFPA to be measured while the IRFPA operates in a normal manner. The spectral response measurements made with this system are validated by determining the peak wavelength responsivity of a long wavelength IRFPA, whose spectral response was measured with the system, using various blackbody temperatures. Sample spectral response data are presented along with the responsivities determined. With this spectral response measurement and a blackbody IRFPA measurement capability, complete radiometric characterization of IRFPAs can be performed.

For over 40 years, NRaD (formerly NOSC) has been involved in radiometric testing and evaluation of long wave infrared optical filters and materials for a great number of military and civilian programs. Over this time, NRaD's cryogenic filter and materials measurement capabilities have evolved to include spectral emittance from 2.5 to 25 microns and spectral transmittance from 1.5 to 50 microns with spectral attenuation measurements to 10^-7. Recent upgrades allow continuous sample temperature control from 4.2 to 300 Kelvins for all emittance and transmittance testing. In addition to radiometric testing, NRaD routinely conducts space environment dose simulation testing on LWIR optical components under sensor operating conditions. Accumulated dose levels from 0.001 to 1 megarad (Si) are easily achieved. Results of recent emittance, transmittance and low level dose testing are included. Measurement theory and data limitations are also discussed.

This research presents the performance evaluation of two retina-like preprocessors for imaging detector arrays. In particular, the preprocessors are an adaptive neural network gain and level controller and a temporal high pass filter. These preprocessors were developed by Scribner et. al. at the Naval Research Laboratory for the purpose of chip-level non uniformity correction associated with InfraRed Focal Plane Arrays. Three performance evaluations are described. The first evaluation involves the determination of a spatio-temporal steady state transfer function (very similar to Modulation Transfer Function). The second involves the processor reduction of fixed pattern noise. Finally, the processor is evaluated on its capacity to reduce temporal l/f noise. The performance of the preprocessors are compared and contrasted with a discussion of their advantages and disadvantages.

A variety of projects have been recently completed or are underway that utilize 3D architectures to achieve major enhancements of focal plane array signal processing capabilities. Progress will be presented in the areas of non-uniformity correction analog-to- digital conversion, spatial and temporal filtering, foveal vision, event-driven multiplexing, and neural pattern recognition. This progress is the result of on-going collaborative and individual efforts under direction of the authors.