'Does it look like a real sensor image?' This is the question that is often asked by users of the TTIM model. A common way to examine the resemblance of simulated and measured images is to compare the MRTD curves obtained from simulation and measurements. But a good agreement between the MRTD curves does not necessarily ensure that the simulated images look like the actual sensor images of real scenarios. In this work, the approach taken to validate the sensor model was to compare images. Thermal images from various scenes and ranges were recorded in the field with known sensors. At the same time, high resolution radiometric maps of the same scenes were measured as well, and later on used as input for the TTIM sensor model. The simulated images were compared to the digitized images obtained from the sensors. This paper describes the field test procedure and presents the results of this comparison. Some of the sensor effects that are not taken into account in the model are identified.

PCTTIM was developed under the joint sponsorship of TARDEC and the 7th Army Training Command in Germany as an instructional tool for the purpose of familiarizing thermal sight operators with a variety of vehicle types, thermal viewer types, and atmospheric effects. For this initial version the design goals were modest. We needed to present a user-friendly interface which allowed the operator to view a thermal image adjusted for atmospheric, optical, detector, and electronic effects. To this end, we took knowledge gained from implementing TTIM under Unix and built a simplified version on an Intel based PC system. PCTTIM is written in C++ (Borland C++ version 3.1 for Windows) and using Borland's Object Windows Library (OWL). This paper is divided into two major sections, a Model Description section and a Future Enhancements section. Each section is subdivided into user interface related issues and IR effects modeling issues. Under the Model Description section, the user interface sub-section is further subdivided by point of view. The student user's perspective is covered first, then the instructior user's perspective is covered.

An image based comparison of modeled IR cameras in the medium- wave (MW) and longwave (LW) bands is done using the TARDEC thermal image model (TTIM) and LOWTRAN7. A state-of-the-art staring focal plane array (FPA), a common module scanning FLIR, and a scanning dual-band sensor are modeled. The simulations using TTIM demonstrate the imaging performance of the cameras as well as the degradation caused by the atmosphere in the two bands. Atmospheric degradation to the image is simulated in rain and fog in northern hemisphere environments.

The current DoD target acquisition models have two primary deficiencies: they use simplistic representations of the vehicle and background signatures, and a highly simplified description of the human observer. The current signature representation often fails for complex signature configurations, yields inaccurate detectability and marginal pay-off predictions for low signature vehicles, is not extensible to false alarms and temporal cues, and precludes vehicle design guidance and diagnosis. The current human observer model is simplified to the same degree as the signature representation, and as such is not extensible to high fidelity signature representations. In answer to the noted deficiencies, we have developed the TARDEC visual model (TVM). We have adopted an alternative approach that is based on emerging academic computational vision models (CVM). Our approach is tailored to visual signatures, though the model is applicable to thermal, SAR as well as other categories of imagery. Color imagery, input to the model, is initially transformed into a 3D color-opponent space comprising luminance, red-green, and yellow- blue axes. Each plane in the color-opponent space is then decomposed by local, oriented spatial frequency analyzers (Gabor or wavelet filters) in keeping with current knowledge of retinal/cortical processing. Signal-to-noise statistics are then calculated on each channel, appropriately aggregated over all channels, and used within the signal detection theory context to predict detection and false alarm probabilities.

Restoration of thermal images distorted by the atmosphere, received on a focal plane array (FPA) thermal imaging system, is presented. The restoration method is based upon atmospheric modulation transfer function (MTF) analysis. Using turbulence and aerosol MTF prediction models atmospheric distortions and image corruption are modeled. Preliminary restoration results indicate significant improvement in image quality.

The target thermal structure is playing a stronger role in modeling the performance of autonomous IR seekers to acquire and track targets. The impact of the target thermal structure on seeker and sensor acquisition has been previously reported. In this paper, the impact of the target's thermal structure on the acquisition and tracking capability of autonomous imaging IR seekers using staring focal plane arrays is assessed. This paper examines both the magnitude of the thermal structure, referred to as the vehicle's thermal standard deviation, and the distribution of the thermal structure, referred to as the power spectral density (PSD). The vehicle's thermal PSD is important in that it determines how much structure the seeker will see given the number of resolvable pixels the seeker has on the target. PSDs that reflect actual armored targets as well as simple warm bodies with a single hot spot are explored. PSDs that are necessary for targets to optimally match the background clutter are also addressed. In addition to the impact of the target's thermal structure, the impact of aliasing effects that can be present in staring seekers is discussed. Relative advantages of trading off resolution versus eliminating aliasing effects are presented.

The staring IR focal plane array (IRFPA), once a novelty, has now evolved into an important technology being used in a variety of imaging systems. These systems frequently include schemes for nonuniformity correction (NUC) that reduce artifacts caused by lack of detector uniformity within the IRFPA. The effect of these artifacts is a fixed pattern that interferes with the desired image of the scene. Although the resulting imagery can be substantially improved using NUC, significant residual fixed pattern noise (RFPN) typically remains. As IRFPAs become more standard in designs for missile seekers, measurement methods related to RFPN should be standardized. This paper introduces an improved method with which the IRFPA fixed pattern noise can be measured both before and after NUC algorithms have been applied. The proposed standard provides two key benefits: (1) consistent quantification of the specific characteristics of the inherent spatial noise or residual post-NUC RFPN that significantly impact system performance; and (2) a standardized process that facilitates IRFPA and NUC comparisons and measures progress toward design improvement. The approach considers the 3D noise modeling technique proposed for FLIR systems and modifies the technique to account for the spatial scale and power spectrum of the RFPN. The resulting characterization is called the simplified power spectrum (SPS). Residual low-frequency nonuniformities such as gradients across the entire array may not be as important as high-frequency, uncorrelated variations; the SPS allows differentiation of these RFPN effects. The algorithms and test procedures required for the standard are presented in the paper.

A new measure called minimum findable temperature difference (MFTD) is offered as a means of characterizing thermal imaging performance under scene clutter limited conditions. MFTD extends an older performance methodology called minimum detectable temperature difference (MDTD) which was offered in the 70's as a sensor performance characterization under noise limited conditions. MFTD departs from MDTD in that it addresses scene clutter plus noise as well as sensor-noise-only operating conditions. In addition, MFTD measures the in-FOV search performance of the thermal imager. This paper details the measurement procedures associated with MFTD using real target vehicles and equivalent bar targets. Methods of characterizing thermal target scenes in terms of vehicle signatures and scene clutter are offered that appear to correlate with measured probability of target finding. Methods used to train MFTD observers and procedures for processing displayed target embedded scenes are provided. Finally, applications of MFTD for modelling manned sensor performance in cluttered environments using TTPF model formulations are offered. A future application of MFTD with real targets to develop standard Army observer ROC curves is suggested as well.

The effect of low frequency mechanical vibrations in the image plane on thermal imaging target acquisition is considered. A model is described that takes as input the mechanical vibration data such as amplitude and frequency and the physical characteristics of the target. The output is the probability of detection as a function of time. Analysis indicates that even low amplitude vibration greatly affects the predicted target detection times and, hence, the utility of the IR system in realistic scenarios.

Recent investigation of atmospheric modulation transfer function (MTF) in the thermal range has shown significant spatial frequency dependence. This dependence is related to aerosol forward scatter as well as absorption effects, shown to be primary contributors to the blur in thermal imaging through the atmosphere. In this paper, the night vision laboratory (CNVEO) target acquisition search model is revised to include atmospheric MTF spatial frequency dependence. It is shown how the target acquisition probabilities and, conversely, the range at which targets can be detected are changed by the inclusion of atmospheric effects. Under weak turbulence conditions target acquisition probabilities are better than predicted by the CNVEO model at shorter ranges. For very strong turbulence, as at midday on hot days, target acquisition probabilities decrease. The improvement possible by correcting for this blur automatically in real time is significant.

The fast objective method for predicting the MTF and MRTD for focal plane array (FPA) systems is presented. The practical method by which the MTF and objective MRTD are obtained is discussed and predictions compared with results from an InSb FPA imager. Applications of the technique for the assessment of both scanned and FPA IR systems are considered.

This paper presents a model for predicting the performance of thermal imaging systems (TIS). This model combines conventional modeling relationships and recently reported characteristics of display monitors to determine the signal-to-noise ratio out of the TIS. Also included are the results of psychophysical experiments which evaluated the capability of a human observer to detect the presence of an object displayed on the same monitor. The model is then used to determine the noise equivalent temperature difference based on background photon noise limited operating conditions of the TIS. Finally, the minimum detectable temperature difference in the scene is determined from the maximum signal-to-noise ratio of the monitor.

The measurement of minimum resolvable temperature difference (MRTD) of a thermal imaging system has always had problems. It is subjective and it requires laboratory conditions (normally air conditioned for a constant temperature), expensive measuring equipment, and a considerable period of time. The problems to be encountered in carrying out this measurement on, for example, the battlefield would be far worse. This paper describes a novel test instrument for quickly making measurements of MRTD in a GO-NO GO, pass or fail, mode in either the quality department of a company or in the field under difficult weather conditions. This paper will also describe the assessment of a demonstrator system and results obtained under laboratory conditions and under conditions where the demonstrator was thermally stressed to mimic field use. This paper will also briefly describe further minor modifications to the equipment which would allow a full MRTD measurement to be made.

This paper will address some of the concerns in the measurement and interpretation of laboratory measured sensor data for use in field performance prediction. The concept of a digital port MRTD will also be discussed with emphasis on its usefulness in the prediction of the man-in-the-loop field performance of the sensor.

This paper presents signal processing techniques used in a new FLIR design to improve the minimum resolvable temperature, to increase the dynamic range, and to provide additional new features. The techniques include analog edge enhancement, low frequency amplitude limiting, digital filtering, automatic gain and level control, and electronic zoom. The system uses an eight- channel SPRITE detector and, as a result, all analog processing must be replicated in each channel. This design uses adequate analog processing to allow the use of eight-bit scan conversion and digital processing without sacrifice of dynamic range. The result is a highly effective tradeoff between function and cost.

A method for microscanning in imaging sensors is developed that allows liquid-crystal beam steerers to be used as nonmechanical microscan devices. This submicroscanning method involves using liquid-crystal beam steerers to shift images on a focal plane array by a fraction of the amount used in typical microscan methods. Interpolation techniques based on interlaced sampling are used to produce images free of aliasing out to twice the Nyquist frequency determined by the focal plane array. Since a continuous phase ramp is produced by the liquid-crystal beam steerer, dispersion effects due to the grating-like nature of the devices are avoided. Simulations for both 1D and 2D cases are presented, as well as experimental results using a 3 to 5 micrometers imaging sensor and a liquid-crystal beam steerer designed for operation at 1.064 micrometers .

The generation of complete databases of IR data is extremely useful for training human observers and testing automatic pattern recognition algorithms. Field data may be used for realism, but require expensive and time-consuming procedures. IR scene simulation methods have emerged as a more economical and efficient alternative for the generation of IR databases. A novel approach to IR target simulation is presented in this paper. Model vehicles at 1:24 scale are used for the simulation of real targets. The temperature profile of the model vehicles is controlled using resistive circuits which are embedded inside the models. The IR target is recorded using an Inframetrics dual channel IR camera system. Using computer processing we place the recorded IR target in a prerecorded background. The advantages of this approach are: (1) the range and 3D target aspect can be controlled by the relative position between the camera and model vehicle; (2) the temperature profile can be controlled by adjusting the power delivered to the resistive circuit; (3) the IR sensor effects are directly incorporated in the recording process, because the real sensor is used; (4) the recorded target can embedded in various types of backgrounds recorded under different weather conditions, times of day etc. The effectiveness of this approach is demonstrated by generating an IR database of three vehicles which is used to train a back propagation neural network. The neural network is capable of classifying vehicle type, vehicle aspect, and relative temperature with a high degree of accuracy.

We describe signature and sensor modeling developments associated with the target identification sensor performance (TISP) imagery- based targeting trainer. The purpose of TISP is to train with a personal computer (PC), IR ground vehicle identification as well as IR and visible imaging principles and the impact of these principles on targeting performance. The generic TISP concept is illustrated as it applies to the new night targeting system (NTS) Laser Range Finder/Designator and IR and visible imaging sensors supporting TOW and HELLFIRE operations on the Marine Corps AH-1W Cobra. Ground vehicle identification skills are developed in TISP using real IR imagery with identification cues. TISP trains FLIR environmental and sensor effects as well as IR principles using real IR imagery coupled with graphical tutorials. Novel algorithms are applied to modify imagery in response to user control over environmental conditions such as weather and time of day, and sensor settings, such as gain, level, field of view (FOV) and polarity. Environmentally dependent range performance predictions for tactical targeting are interactively displayed, also using real imagery coupled with graphics. TISP contains imagery of sixty ground vehicles and five backgrounds.

System-level performance testing of IR focal plane arrays (FPAs) at low background flux has been demonstrated prior to system- level integration using an optical simulator designed at the Hughes Technology Center. The optical simulator uses 30K cooled reflective optics and scanner to focus a 55 micrometers diameter spot on the array under test. The scanner can be synchronized with the array's on-chip time delay and integration (TDI) functions to simulate system-level FPA performance. The optical simulator design incorporates a separately detachable dewar to allow the device under test to be changed without warming up the optics. The optical simulator was designed to test devices in the 3 micrometers to 30 micrometers region and contains cooled background and signal wheels which allow for testing at multiple wavelengths and background flux levels. This paper describes the optical simulator and presents the initial test data obtained.

This paper describes a method of modeling and testing of a modular imaging spectrometer instrument (MISI), with special emphasis on system and sub-system modulation transfer function (MTF) analysis. The optical system was modeled using optical ray tracing methods. The dynamic deformation of the scan mirror was modeled using a finite element analysis method, and the image degradation due to the deformation is estimated using optical image formation theory. The detector and conditioning electronics were also modeled using the transfer function theory. This modeling approach was used as a tradeoff tool for the design of MISI. Laboratory experiments were conducted to test the performances of each sub-system on design criteria, and finally a field test is planned to test the overall optical/mechanical/electrical performance of the entire imaging chain.

In recent years, with the advance in analytical and calculating ability of scientific spreadsheets and the increasing speed of PC machines, it is now possible for any sensor system designer to carry out basic performance analysis without always having to refer to special purpose performance models. The designer can use these spreadsheets to custom-make any system and subsystem models for his own purpose. The spreadsheet models presented in this paper will demonstrate the analysis of horizontal and vertical 4- bar target responses of a serial scan FLIR system by using a commercially available scientific spreadsheet (i.e., MathCAD version 4.0). The results from the model have shown excellent correlation with actual measured data. Finally, this ability to accurately predict subsystem performances have proven to be extremely useful not only during the design phase, but also during the prototyping and preproduction integration phase. Because during trouble-shooting efforts, it provides the engineers the ability to understand subsystem problems in an almost real time fashion.

We present a new model for predicting the minimum resolvable temperature difference (MRTD) curve of thermal imaging systems. The analysis for the new model concentrates on contrast reduction due to spatial frequency limiting factors of subsystem components. Curves have been generated for this model for a system with typical component values. These results are compared with curves generated from the NVL's static performance model. The proposed visibility model leads to a relatively simpler development for a MRTD predictor which can readily account for artifacts due to a nonzero system phase transfer function. In addition the visibility model makes no assumptions regarding the recognition process and therefore is adaptable to the goal of modeling an objective MRTD measurement. The visibility model agrees with the static performance model except at very low and very high spatial frequencies where the proposed model appears to be in better agreement with observed trends in measured MRTDs.

The Arnold Engineering Development Center (AEDC) has developed new test technologies and methodologies for realistic mission simulations testing of IR space-based sensors. These technologies and methodologies have been combined into an integrated approach for space sensor testing. Direct write scene generation plays a critical role in this integrated approach and is being applied in two new AEDC test facilities. Prior to performing the first test in such a new facility, a critical but often overlooked process must be completed. This critical process demonstrates that the test facility can indeed provide a realistic, NIST traceable simulation of a sensor's mission. This process is complex and must be uniquely tailored for each individual test facility and sensor mission. Such a process can be designed to address both developmental test and evaluation and operational test and evaluation concerns. A case study based on AEDC's direct write scene generation technology will be used to illustrate the issues related to the validation, verification, and accreditation process.

The automated minimum resolvable temperature (AMRT) test method for IR sensor performance testing is developing into an accepted method of automated test. This paper gives the basis of testing AMRT and shows how it relates to the manual measurement that has been correlated to field performance. The lessons learned as a result of implementing AMRT at Northrop are reported in this paper. These lessons are discussed in the areas of developing algorithms and methodologies which work well in the typical noise environment. Improved methods have been developed to give a better measure of signal response in the presence of many kinds of noise. Various types of noise measurements are addressed as well as their impact on the resulting AMRT.

We present the first astronomical results from the new 3D near IR imaging array spectrometer. These include K band (1.95 to 2.45 micrometers ) spectra and images of nearby starburst galaxies and active galactic nuclei with a spectral resolution of 1000. A special image slicer allows simultaneous spectra and imaging of an 8 arc second field of view. The background limited performance achieved by this instrument represents an order of magnitude reduction in integration time over existing near IR cameras and spectrometers. In addition, subtraction of atmospheric OH lines may be performed with far higher accuracy. We discuss the data reconstruction procedure, with special emphasis on flat fielding and calibration of the detector. This is complicated by the scrambled image format, which results in adjacent image pixels being widely separated on the detector. Small non linearities of the optical elements must also be dealt with carefully. We also discuss future improvements to instrument performance, including a low order adaptive optics system for compensating atmospheric turbulence.

Magnavox EOS has designed and prototyped two compact Stirling- cycle cryocoolers, the MX7050 and the MX7055, for IR imaging applications. The MX7050 and the MX7055 are designed to meet respectively the U.S. Army B2 specification for the 0.35 watt linear cooler and the 0.6 watt standard advanced dewar assembly IIIB linear cooler. The design objectives were achieved with a common compressor and two different cold expander designs. The MX7050 and the MX7055 prototypes demonstrated cooling capacities far exceeding the design specification over the ambient temperature range of -54 degree(s)C to 71 degree(s)C. Test results showed that efficient cooling up to 1.2 watt at 77K was obtained with a MX7055 prototype. The compressor utilizes a dual opposed piston configuration to achieve low output vibration. The coolers are designed to exceed 4000-hour life and weigh about 2 pounds without driver electronics. Cooler electronics for DC-AC conversion and temperature control are available both as an internal part or external part to the cooler. This paper describes the design, fabrication, performance, and test results of the Magnavox MX7050 and MX7055 prototypes.

This paper presents the measurement requirements and algorithms that characterize forward looking IR (FLIR) imaging systems in the Martin Marietta EO Characterization Lab. The algorithms presented automate the following major measurement requirements: signal transfer functions, 3D noise, noise equivalent temperature difference, fixed pattern noise, nonuniformity, modulation transfer function, and minimum resolvable temperature difference.

A model is proposed for determining the effect of image clutter on the time-line behavior of observer eye fixations when searching for targets in IR scenery. Several possible metrics for qualitatively evaluating the degree of clutter are tested. Correlations between these metrics, the probability of fixating a particular area and the length of time fixating a particular area will be discussed.