The Minimum Resolvable Temperature Difference (MRTD or MRT) is the most widely accepted and inclusive figure of merit for describing a thermal imaging system's performance. It is the product of analytic mathematical models and traditional man-in-loop system hardware performance measurements that describe IR systems. MRT is a basis for thermal field performance model predictions and is commonly used in specification of thermal imagers. The MRT test is subjective because it requires human observers to just discern increasingly smaller 4-bar patterns as a function of temperature differences between bars and the background. When performed by trained observers, the MRT test is an accurate measure of sensitivity as a function of spatial resolution. The ability to resolve 4-bar patterns varies between observers. Furthermore, MRT is a psychophysical task, for which biases are unavoidable. In this paper, uncertainties in MRT measurements are reported for individual trained observers and between observers as functions of some biases, such as random and fixed pattern noise. For this paper, virtual MRTs were performed on a new, custom visual acuity test simulator, developed for NVESD, that allows precise control over significant sensor and display parameters, and these results are compared. Through a process of eliminating sources of MRT variability, we have been able to quantify the observer variability.
A new method for obtaining the geometric resolution of infrared cameras with focal plane array (FPA) detectors is
introduced. Selected values of the modulation transfer function (MTF) are measured in real time on a frame-to-frame
basis. For this purpose a test bar pattern is imaged onto the FPA plane so that the bar image has a spatial frequency
which is detuned with respect to the FPA Nyquist frequency. This generates a beat signal whose amplitude corresponds
to the contrast of the bar image to be measured, without the sampling MTF. If the test bar pattern covers a range of at
least half a beat period the modulation amplitude can be determined independent of the actual phase position. This
method can be adapted to real-time MTF measurements. Examples are presented for shaker excitations of an IR camera
on a platform with means for active and passive stabilization. Sweeping the frequency in sinusoidal excitation mode
exhibits resonance structures deteriorating the camera resolution thus giving valuable information for optimizing the
design of camera and suspension system. Residual stabilization errors are measured under random vibration excitations
for each individual frame, their distribution being conveniantly visualized by bar histograms.
The performance of modern Forward-Looking Infrared (FLIR) systems is vital to preserving the lives of military personnel and civilians. Proper test and evaluation processes enhance system integration and provide useful data for system improvements and tactical implementation. Such processes are also necessary to effectively compare modeling and laboratory data with in-situ flight data. In this paper we discuss the test and evaluation processes typically used for infrared imaging systems and discuss some methods of correlating laboratory and modeling results with flight data. We will present a hypothetical infrared imaging system, demonstrate how it is put through various stages of testing, observe sample hypothetical system data results, and discuss some reasons for selecting specific testing facilities and locations. Additional discussion will focus on preliminary testing considerations such as selection of ground targets for Minimum Resolvable Temperature Difference tests and the selection of proper target FLIR test patterns. Finally we will mention future testing challenges such as short-wave infrared systems and the validity of current assumptions used when testing mid-wave infrared systems.
BAE SYSTEMS produces hundreds of low cost, high performance, uncooled IR imagers each month for use in commercial and military applications. The production process of each imager includes several steps that begin at the wafer level and end at an in-camera test. Each step is critical to end yield improvement by detecting failure at various stages in the production flow. Both automated test equipment and an integrated database system are essential at each phase to efficiently build and automatically configure cameras for each customer. This paper discusses the process and tools used to reliably test and ship uncooled thermal imagers in addition to specific methods and calculation techniques for characterizing key performance parameters such as Responsivity, Noise Equivalent Temperature Difference, and Operability.
This paper presents the latest developments in instrumentation for military laser range-finder/designator (LRF/D) test and evaluation. Santa Barbara Infrared (SBIR) has completed development of a new integrated laser test module supporting a wide range of laser measurements including range accuracy and receiver sensitivity, pulse energy and temporal characteristics, beam spatial/angular characteristics, and VIS/IR to laser co-boresighting. The new Active Laser Test Asset (ALTA) incorporates all the functionality of the previous Active Range Module (ARM) and Laser Test Module (LTM) in a form factor suited to both modular/portable EO test systems and standard product configurations. Key discriminators of the ALTA design include a three-color, fiber-coupled laser source (1064, 1540, and 1570 nm), a simplified optical path design, and enhanced laser output energy density in all three wavebands.
In traditional designs of a NUC (non-uniformity correction) system, a rotating chopper-wheel (or a blurring/deform lens) is used to separate the outside scene and the inside FPN (fixed pattern noise) on the FPA (focal plane array). To design a NUC system removing the chopper-wheel (chopper-free) and its control electronics and hardware will not only considerably reduce the cost, but also require less space to fit the NUC system. In this paper, we describe a recently developed CF (chopper-free) NUC system. This system is simpler to build, costs less, and requires less space, as compared with traditional designs.
By jointly optimizing the design of optics, mechanics, and electronics systems with reduced size, weight, and cost can be realized. This joint optimization acts to increase the system trade-space compared to systems that optimize each component separately. Increasing the size of the system trade-space allows highly customized system design. An example of joint optimization is given for a LWIR imaging system with a conformal first surface. This example demonstrates an approximately 50% reduction in size, weight, and cost compared to acceptable traditional system solutions.
Santa Barbara Infrared, Inc (SBIR) has developed a dual-band infrared spectroradiometer for highly accurate radiometric calibration of electro-optical (EO) test stations, light sources, and optical surfaces. The "RAD-9000" design covers the 3-5 mm and 8-12 mm spectral bands, provides thermal sensitivity of better than 40 mK, supports object temperatures from 278-373 K, and delivers better than 2% spectral resolution (Dl/l). The RAD-9000 features computer-controlled operation, an intuitive graphical user interface (GUI), motorized focus adjustment, VIS-CCD sighting/alignment capability, less than 2 mrad detector IFOV, and an internal ambient reference for background subtraction and enhanced stability.
In addition to high-performance relative radiometry, the RAD-9000 offers a high degree of absolute radiometric accuracy by utilizing a dedicated radiometric reference module. The reference module incorporates two 8-inch, variable temperature, high-emissivity extended sources to provide a stable, accurate absolute radiometric reference external to the main optics.
Indigo Systems, a division of FLIR Systems, Inc., has released a commercial off-the-shelf, PC-based software program named RPro. RPro, an optional component of Indigo's RTools Radiometric Software Toolkit, was developed for engineers and scientists to efficiently batch process and analyze data from high-end infrared focal plane array cameras, Fourier transform infrared (FTIR) spectrometers, high-speed radiometers, and imaging spectrometers. Many core radiometric calibration and data reduction algorithms already exist within RPro for the user with minimal infrared radiometry experience. For the advanced radiometry engineer, RPro provides a flexible and extensible graphical programming interface to easily develop custom radiometric calibration and data reduction algorithms. Moreover, adding the RPro component to RTools provides the radiometry engineer with the capability to quickly create a data reduction algorithm that when used in conjunction with MODTRAN will correct for atmospheric effects using range supplied time, space, and position information (TSPI). RPro was designed to integrate seamlessly with all other RTools components by utilizing the Standard Archive File (SAF) format maintained by the U.S. Air Force at Arnold Air Force Base, TN.
We present a new measure called target identifiability, as an efficient alternative for measuring identification scores. Identifiability is operationally defined as the amount of blur required to reduce the target signature to its identification threshold. It can quickly be determined using a simple adjustment procedure. To validate the new measure, we measured the identifiability of targets in a set of real and simulated thermal images. The identification scores for these targets were available from a previous study. Our results show that identifiability indeed determines identification performance. Sufficient accuracy can be obtained with only a few (typically 2 or 3) trained observers. The associated measurement procedure is simple and requires only a limited amount of time.
We investigate the effect of band-limited masking noise and blur on the minimum contrast required to identify a) military targets, b) the standard MRTD and MTDP four-bar test pattern and c) the TOD equilateral triangle test pattern. First, the image containing the test object was spatially blurred by an amount varying from one pixel up to the maximum blur at which the object could still be identified or resolved. This mimics the effect of sensor optics and detector blur on the image of a target at different ranges. Then, band-limited noise was added to the image. The center spatial frequency fc of the masking noise was varied over 7.0 octaves. Observers had to indicate at which target contrast they were just able to identify the target. The results are a) identification thresholds for all targets are strongly elevated by masking noise of certain spatial wavelengths and much less by coarser or finer noise, showing that identification by human observers is mediated by a relatively narrow visual filter, and b) with increasing blur, maximum masking shifts towards lower noise spatial frequencies in a consistent but non-linear way. Current popular TA models are inconsistent with these results but suggestions for improvement are presented. The laboratory test patterns are appropriate to characterize target acquisition performance with viewing systems that include noise and blur.
This paper describes research on the determination of the fifty-percent probability of identification cycle criterion (N50) for two sets of handheld objects. The first set consists of 12 objects which are commonly held in a single hand. The second set consists of 10 objects commonly held in both hands. These sets consist of not only typical civilian handheld objects but also objects that are potentially lethal. A pistol, a cell phone, a rocket propelled grenade (RPG) launcher, and a broom are examples of the objects in these sets. The discrimination of these objects is an inherent part of homeland security, force protection, and also general population security.
Objects were imaged from each set in the visible and mid-wave infrared (MWIR) spectrum. Various levels of blur are then applied to these images. These blurred images were then used in a forced choice perception experiment. Results were analyzed as a function of blur level and target size to give identification probability as a function of resolvable cycles on target. These results are applicable to handheld object target acquisition estimates for visible imaging systems and MWIR systems. This research provides guidance in the design and analysis of electro-optical systems and forward-looking infrared (FLIR) systems for use in homeland security, force protection, and also general population security.
Imagers based on focal plane arrays (FPA) risk introducing in-band and out-of-band spurious response, or aliasing, due to undersampling. This can make high-level discrimination tasks such as recognition and identification much more difficult. To overcome this problem, three-chip color charge coupled device (CCD) cameras typically offset one CCD by 1/2 pixel with respect to the other two. Analogously, monochrome imagers including infrared can use microscan (or dither) to reduce aliasing. This paper describes a generic microscan technique and benefits of microscanning. Covered are analysis and experiments on four-point microscan employed in infrared imagers, in which the image is mechanically shifted by 1/2 pixel between fields, in each dimension. Four of these offset fields are then combined to form one frame of high-resolution video. We show that microscan reduces aliasing, which results in higher resolution and improved image quality resulting in improved performance.
The enhancement of undersampled imager performance has been demonstrated using super-resolution techniques. In these techniques, the optical flow of the scene or the relative sub-pixel shifts between frames is calculated and a high resolution grid is populated with spatial data based as a result of scene motion. Performance enhancement has been demonstrated for the case of a static image with the undersampled imager output compared to a static image that has been acquired through a frame series in a dynamic scene. In this research, the performance is compared for four cases: static image with undersampled imager, static image with super-resolution frame sequence, dynamic image with undersampled imager, and dynamic image with super-resolution frame sequence.
The spatial resolution of Focal Plane Arrays (FPA) is affected by sampling. The artifacts introduced by the sampling procedure are usually referred as "aliasing". Phase artifacts are introduced by the non-isoplanatic nature of the sampling mechanism in FPA. In this paper we introduce a stochastic description of these artifacts. This approach allows us to elucidate similarities and differences between "aliasing" and "phase effects". Figures of merit are introduced in order to characterize regions of isoplanatism in the Fourier Space. The relation between these figures of merit and target perception models is explored in order to clarify further research.
Estimates of the performance of an imager working in a turbulent atmosphere can often be obtained if an estimate of the index of refraction structure constant Cn2 is available. In this paper, results from predictive parametric models for are compared with long-term measurements of Cn2. A brief overview of imaging in turbulence is given. The impact of turbulence on imager performance and models to account for it are reviewed. This is followed by a presentation of the methods used to collect Cn2 data. Some published parametric models for predicting based on meteorological data are reviewed. These models are then compared with Cn2 measurements taken over several years. Performance of the models as a function of time of day and season of year are shown. General conclusions regarding the utility of using these models to predict system performance are presented.
This paper describes a simulation technology for HgCdTe infrared detectors used in advanced IR focal plane array architectures. This model addresses the material processes needed for fabrication and the electrical characteristics of multi-layer structures covering a wide range of wavelengths from middle wavelength to very long wavelength.
To simulate an Enhanced Vision System (EVS), CEA/LETI Infrared Laboratory has developed two behavioural models of infrared focal plane arrays : one in the Short Wave IR and the other in the Long Wave IR band. These Infrared Focal Plane Arrays (IRFPAs) models will be implemented on simulation platform aimed at evaluating the impact and use of infrared sensors in automotive and aeronautic applications. To be realistic, model parameters are extracted from electro-optical characterization of real components. The SWIR detector is calibrated with a 320x256 HgCdTe cooled FPA component from SOFRADIR, and the LWIR one with an uncooled micro-bolometer array from ULIS (a_Si technology from LETI). The flexibility of the models allows to simulate cameras based on these components and to forecast future ones based on different read-out circuit or detector technologies.
In this paper we present the IRFPAs models, the main electro-optical characterization results and we compare some experimental measurements with simulations.
Active imaging systems with coherent or partially coherent illuminators differ significantly from passive incoherent imagers. The impact of atmospheric turbulence in active systems can be greater than in a passive systems and rough surface induced speckle is often present in the imagery. The nature of the target illumination is also of a significantly different character than in passive systems. Currently released NVESD models do not address these aspects of imaging for full or partially coherent illumination. Extensions of the NVESD models to account for the performance impact of turbulence and speckle have been completed. This paper outlines the methods used for modeling the performance of active imaging systems with particular attention to the impact of speckle and atmospheric scintillation. Results from perception experiments using simulated imagery are presented.
Laser Range-Gated (LRG) imagers provide high contrast images of targets at extended ranges because the laser light scattered by the intervening atmosphere is gated out. Atmospheric backscatter of the laser light does not degrade the contrast of LRG imagery. Natural illumination helps range performance by increasing overall target illumination. However, natural illumination of the intervening path degrades target to background contrast, and this hurts range performance. This paper provides a model for predicting the influence of natural illumination on LRG performance.
This research compares target detection in the longwave and midwave spectral bands in urban environments. The Night Vision and Electronic Sensors Directorate (NVESD) imaged one hundred scenes at several Army Military Operations in the Urban Terrain (MOUT) sites during day and night. Images were resized to make the field-of-view (FOV) for each scene approximately the same. These images were then presented in a time-limited search perception experiment using military observers. Probabilities of detection were compared between the two spectral bands. Results from MOUT search were compared with previous modeling efforts.
The miniaturization of light detectors in the visible and infrared has produced devices with micrometric and sub-micrometric spatial features. Some of these spatial features are closely linked with the physical mechanism of detection. An example of these devices is an optical antennas. To spatially characterize optical antennas it is necessary to scan a probe beam on the plane of the optical antenna. The mapping of this response is then treated and analyzed. When the response of the antenna is monitorized at visible or near-infrared frequencies, a sub-micron scanning step is necessary. In this paper we show the experimental set-up of a measurement station having a spatial resolution of 50 nanometers. This station is devoted to spatially characterize micrometric detectors, and specially optical antennas. The origin of the uncertainties of the measurement protocol is shown and practically analyzed. This station is also applied for characterizing the temporal, spectral, and polarization sensitivity specifications of light detectors with the previously mentioned resolution.
Humans cannot objectively judge electro-optical imaging systems looking on an image of typical scenery. Quality
of the image can be bad for some people but good for others and therefore objective test methods and advanced
equipment are needed to evaluate these imaging systems. Test methods and measuring systems that enable reliable
testing and evaluation of modern thermal cameras, color and monochrome TV cameras, LLLTV cameras and image
intensifier systems are presented in this paper.
This paper analyzes the various factors contributing to the performances of the FLIR, and describes a series of solutions to control its transfer function and improve its performance. The paper describes the camera concept, elaborating on its behavior and associated control functions in the time domain, on the two dimensional image domain, and on the signal model translation to standard video signal. The camera contains the following blocks :
A. Time domain low pass filter that is controlled in real time according to the velocity of the objects observed
B. Two dimensional high pass filter that is controlled in real time according to SNR of the input video signal,
C. Supplementary filter that removes the 1/f noise, controlled according to the SNR of the input video signal,
D. An automatic controlled dynamic range compression mechanism.
Thermal imaging cameras are rapidly becoming integral equipment for first responders for use in structure fires. Currently there are no standardized test methods or performance metrics available to the users or manufacturers of these instruments. The Building and Fire Research Laboratory (BFRL) at the National Institute of Standards and Technology (NIST) is developing a testing facility and methods to evaluate the performance of thermal imagers used by fire fighters to search for victims and hot spots in burning structures. The facility will test the performance of currently available imagers and advanced fire detection systems, as well as serve as a test bed for new technology. An evaluation of the performance of different thermal imaging detector technologies under field conditions is also underway. Results of this project will provide a quantifiable physical and scientific basis upon which industry standards for imaging performance, testing protocols and reporting practices related to the performance of thermal imaging cameras can be developed. The background and approach that shape the evaluation procedure for the thermal imagers are the primary focus of this paper.
The Building and Fire Research Laboratory (BFRL) at the National Institute of Standards and Technology (NIST) is developing a new bench-scale testing facility and methods to evaluate the performance of thermal imagers used by fire fighters to search for victims and hot spots in burning structures. A larger-scale laboratory testing facility was constructed in 2002. This facility was used to determine the effects of water sprays on the imaging performance of a selection of thermal imagers. A new, smaller-scale laboratory facility, currently under construction, will provide a carefully controlled laboratory setting in which aspects of the environment inside a burning structure are simulated as closely as possible. It will also serve as a test bed for new technology. An evaluation of the performance of different thermal imaging detector technologies under field conditions is also underway. Results of this project will provide a quantifiable physical and scientific basis upon which industry standards for imaging performance, testing protocols and reporting practices related to the performance of thermal imaging cameras can be developed. In this paper a description of the testing facilities, including both generations of laboratory apparatus is presented.
This paper describes the use of a rotating test pattern or reticle to measure the Modulation Transfer Function (MTF) of a staring array sensor. The method finds the Edge Spread Function (ESF) from which the MTF can be calculated. The rotating reticle method of finding the ESF of a sensor has several advantages over the static tilted edge method. The need for precise edge alignment is removed. Motion blur is used to simultaneously average out the effect of undersampling and to oversample the edge. The improved oversampling allows reduction of the noise in the generated ESF while keeping a high resolution. A unique data readout technique reads edge data perpendicular to the edge. Perpendicular readout eliminates the need to know or estimate the slope of the tilted edge. This MTF measurement method is validated using simulation and actual data captured by a digital camera. The resulting ESF plots agree well with expected results.