Radiometric infrared camera systems are often used at test ranges to characterize the IR signature of targets such as aircraft or rockets through significant air columns that reduce the received signal through a combination of absorption and scattering. The dominant effect in clear air is molecular resonant absorption which is particularly strong in the midwave IR band (3-5 microns) for carbon dioxide and water vapor. Tactical targets can be imaged at standoff distances up to 1000km or more, but there are many cases where these targets are within a 1km range, as is the case with a close-in flyby at a test range. Therefore it is useful to model the short-range atmospheric transmission to predict its effect on radiometric measurements. Many industrial processes that occur in large outdoor facilities also lend themselves to radiometric measurement for standoff ranges of tens or hundreds of meters. This paper compares experimental radiometric data taken at ranges under 1km to a theoretical model of the atmosphere, and describes a simple method for correcting for air column effects at these relatively short ranges. The data were collected in the 3-5 micron band using an indium antimonide staring-array camera and a long focal length lens combined with radiometric analysis software. The system was calibrated to measure target radiances, but can also be used to estimate target temperatures in cases where the in-band emissivity of the target is well understood. The radiometric data are compared to a model built on MODTRAN code, with conclusions about the attenuation introduced by the atmosphere for standard medium-range imaging systems in "typical" observing conditions. Effects caused by the MTF of the lens system are studied briefly, and used to set limits on the minimum number of pixels the target can subtend and still have an accurately measurable radiance.
The estimation of the performance of electro-optical systems depends on the accuracy of the atmospheric models being used in the propagation prediction codes. In the present work we demonstrate that in a real atmosphere (above the surface layer) the turbulent field of passive scalar fluctuations can differ from Kolmogorov's model. From the spectrum of the intensity fluctuations of LIDAR signals scattered by aerosol concentration inhomogeneities, the behavior of atmospheric turbulence spectrum (power law exponent γ) is estimated. As follows from the experimental data the power law exponent of the turbulent spectra can be different from the case of purely Kolmogorov (γ=5/3). Also, results of lidar measurements of aerosol size distribution, volume, and number concentration at different heights in the Mediterranean region (Be'er-Sheva, Israel) and comparison with models (AFGL, MODTRAN) are presented. Implications can be important for optical communication, imaging through the atmosphere, and adaptive optics.
In this paper, beam scintillation effects on the illumination of targets are quantified with respect to human observer performance. The theory of man-in-the-loop target acquisition is reviewed along with the principles of operation of an actively illuminated gated incoherent imager. A model to account for the effects of beam scintillation on target acquisition performance is introduced. A simulation of beam propagation was developed and used to apply a nonuniform illuminating field to a series of monochrome images of targets. These images were then used in a perception experiment with trained observers. The results of these perception experiments compare well with predictions from the model.
There is strong evidence to suggest that polarimetric techniques offer significant improvements in the ability of electro-optic sensors to detect difficult targets in cluttered backgrounds. Many previous attempts to quantify the potential benefits have been hampered by an inability to gather all the polarimetric data simultaneously from a scene. Sequential data gathering can lead to artefacts in the polarimetric data which in turn lead to spurious and erroneous conclusions being drawn This paper describes work undertaken to build and test a pair of four camera, real time, polarimetric sensors that measure all four Stokes parameters simultaneously. One of the sensors operates in the visible waveband and the other in the near infrared. Example images obtained with both sensors are shown, together with measured target and background signal distributions for one of them. Preliminary results from this work show that the sensor can significantly improve target discrimination.
A new firing post MILAN ADT ("Advanced Technology") is developed by EADS-Lenkflugkoerpersysteme GmbH with the aim to improve the performance of the MILAN weapon system substantially while maintaining all operational features to which MILAN operators are accustomed. The missile tracking sensor of MILAN ADT is now equipped with a single, wide field-of-view optics and a large CMOS detector covering both gathering and guidance phase. Using adaptive windowing and sub-sampling functions of the detector combined with differential imaging modes, all types of MILAN missile are localized with optimum precision over the entire flight path. Another novel feature is the integration of a thermal imager into the optical scheme of the MILAN ADT guidance unit. This replaces the earlier ancillary TIs MIRA and MILIS thus saving the weight of the additional housing and reducing logistic effort. The TI image is displayed on an internal micro-monitor and projected into the eyepiece of the daysight. Optimum boresight harmonization between both missile tracking and sighting channels is ensured by projection of reference marks into each optical sensor path from a common multispectral projector. MILAN ADT is compatible with all existing MILAN missile versions and with MIRA and MILIS TIs; the integrated TI is offered as an option. A planned future range increase of the MILAN weapon system will also be discussed in brief.
A Naval Infrared Search and Track (IRST) demonstrator has been developed for the UK Ministry of Defence. The system uses two staring infrared cameras and split field of view optics to provide panoramic surveillance of the horizon. Use of staring detectors provides improved sensitivity and faster update rate than current scanning IRST systems. The demonstrator is fitted with commercial 640x512 pixel medium waveband detectors but is designed to accommodate 1024x768 pixel detectors in the future. The system features switchable spectral filters to allow choice of the optimum waveband for the prevailing environmental conditions and beam steering optics for non uniformity correction and image stabilisation. Real time processing has been implemented using a combination of Field Programmable Gate Array and PowerPC hardware for detection and tracking. The paper describes the system and presents some examples of its output.
A new laser beam rider guidance system was developed for hypervelocity missile. The technique is based on an off-axis laser beam that is pointing at a constant offset of the target. The missile's sensors get the guiding information by detecting the scattering of the beam created by the particles present in the atmosphere. In order to evaluate the performances of the technique, an hardware in the loop facility was implemented. Three degrees of freedom duplicating the vertical, lateral and spin axes of the missile are represented and the 6 DOFs are modeled. The features of the guidance technique as well as the details of the hardware in the loop facilities are presented.
As thermal imaging technology matures and ownership costs decrease, there is a trend to equip a greater proportion of airborne surveillance vehicles used by security and defence forces with both visible band and thermal infrared cameras. These cameras are used for tracking vehicles on the ground, to aid in pursuit of villains in vehicles and on foot, while also assisting in the direction and co-ordination of emergency service vehicles as the occasion arises. These functions rely on unambiguous identification of police and the other emergency service vehicles. In the visible band this is achieved by dark markings with high contrast (light) backgrounds on the roof of vehicles. When there is no ambient lighting, for example at night, thermal imaging is used to track both vehicles and people. In the thermal IR, the visible markings are not obvious. At the wavelength thermal imagers operate, either 3-5 microns or 8-12 microns, the dark and light coloured materials have similar low reflectivity. To maximise the usefulness of IR airborne surveillance, a method of passively and unobtrusively marking vehicles concurrently in the visible and thermal infrared is needed. In this paper we discuss the design, application and operation of some vehicle and personnel marking materials and show airborne IR and visible imagery of materials in use.
Active imaging systems allow obtaining data in more than two dimensions. In addition to the spatial information, these systems are able to provide the intensity distribution of one scene. From this data channel a certain number of physic magnitudes that show some features of the illuminated surface can be recovered. The different behaviours of the scene elements about the directionality of the optical radiation, wavelength or polarization improve the ability to discriminate them. In this work, the capabilities of one 3D imaging laser scanner have been tested from both dimensional and radiometric points of view. To do this, a simple model of the observing system and the scene, in which only the directional propagation of the energy is taken into account, has been developed. Selected parameters corresponding to transmission, reception and optomechanical components of the active imaging system describe the full sensor. The surfaces of a non-complex scene have been divided into different elements with a defined geometry and directional reflectance. In order to measure the directional reflectance of several materials in the specific wavelength where the laser scanner works, a laboratory bench has been developed. The calculation of the received signal by the sensor has been carried out using several radiative transfer models. These models were validated by experiments in a laboratory with controlled conditions of illumination and reflectance. To do this, a certain number of images (angle, angle, range and intensity) were acquired by a commercial laser scanner using several standard targets calibrated in geometry and directional reflectance.
Large-mode-area double clad fibers offer excellent efficiency and beam quality, high output power as well as lightweight, robust and reliable packaging. The addition of polarization maintaining property though use of well-know Panda-structure has further increased the interest in double clad fibers, especially in fibers doped with ytterbium (Yb). Many material processing, military and R&D applications benefit from wavelength conversion by nonlinear effects, from IR through UV, of 1064nm Q-switched pulses through polarization maintaining large-mode-area double clad Yb-fiber amplifiers. The possibility of power scaling through coherent beam combining has also been identified by the military. The design of a polarization maintaining large modea area double clad fiber for the above mentioned applications must address several key performance parameters: provide large mode area (>300μm2), high efficiency (>80% slope PCE), high average power (>100W), high birefringence (>2*10-4) and offer good beam quality (M2 <1.5), short fiber length (<3m), as well as high reliability and good usability. Further optimization of the fiber design must take into consideration the impairment of the fiber by thermal loading as well as coiling of the fiber for elimination of higher order modes. This paper presents the key design considerations of such fibers for high-average-power pulsed amplifiers and provides the latest experimental techniques to verify the results. The design and results on high performance highly Yb-doped polarization maintaining large mode area fiber manufactured by the Direct Nanoparticle Deposition technology are presented and possibilities and opportunities brought by this technology are discussed.
Our objective has been to find a preferred method for the identification of static targets in single IR images, concentrating on appearance-based methods. This has included thermal modelling of IR signatures and the identification of images of different objects with variation in pose and thermal state. Using principal component analysis, the variances among the images are extracted and represented in a low-dimensional feature eigenspace. Any new image can be projected into the eigenspace by taking an inner product with the basis. The object of interest can be recognized by a nearest-neighbour classification rule, made more accurate by application of over-sampling to the surface manifold by B-spline surface fitting, and made more efficient by a k-d tree search algorithm. To address the problems of recognizing targets in noisy and cluttered images, we have employed a random sampling approach that is based on the principle of high-breakdown point estimation. We have generated a database of images using visible and thermal cameras, in addition to scene simulation software, for use in the learning and recognition/evaluation phases. Our experiments indicate that application of the robust algorithm can reduce the recovery error of the true model image data, for example by a factor of five when the images contain 40% randomly changed image pixels.
This paper addresses the problem of tracking a target in an IR video sequence using a kernel based histogram representation of the target. In this field, gradient ascent methods have demonstrated useful results with weighted kernels and in particular Mean Shift is currently the most commonly used gradient scale method. Our approximation follows the work made by Hager, that uses a SSD objective function (derived from Matusita metric) and combines it with a Newton-like maximization method, resulting a fast gradient scale system. An important property is that this method enables the use of multiple kernels, allowing a more powerful representation with a minimum increasing of computational cost. We analyse the limitation of this representation using the Newton maximization algorithm and we introduce the concept of direction of ambiguity. This concept allows a criterion for choosing the kernels that drive the iteration to minimize the error criterion. The results we present show the improvements of the method over a tracking problem. The target is a small car with a great background similarity.
The most straightforward way to describe the performance of an image intensifier tube, especially under adverse conditions, is to predict the image it yields. In this work we have developed two different methods to provide realistic simulated images in low light level conditions: 1) Approximate Physical Model. A classical approach based on the simulation of the different degradation sources. It provides a good understanding of the image formation process. 2) Synthesis-by-analysis of real images. The observed noise is modelled through texture analysis tools and the image blur through the MTF. The resulting simulated images for both methods were compared with real intensified images (laboratory chart sights and natural images) taken under controlled conditions, close to the performance limits of the image intensifier tube. Both methods generated good results in terms of visual comparison for different object sizes, contrasts or luminances. These methods can be used as a new tool to predict the performance thresholds of the image intensifier. Only well-known or measurable parameters were used as input for the methods.
This paper attempts to inform and educate readers to some of the more recent changes in infrared material pricing and availability along with the impact of optical manufacturing trends and the means by which optical designers, manufacturers and product end-users are using to avoid or eliminate the pitfalls these changes may create.
During many years high volume commercial applications of infrared optics have been slowed down by several cost factors. The development of focal plan arrays and uncooled detectors has allowed to greatly reduce the cost of infrared detectors. In the meantime, Umicore IR Glass has developed an industrial process to manufacture low cost chalcogenide glasses with well controlled properties. These glasses called GASIR 1 and GASIR 2 are transparent in the NEAR and FAR infrared atmospheric windows and are mouldable into high quality finished spherical, aspherical and diffractive lenses. The moulding process allows high volume production of cost effective infrared optics. After the development of several optics in GASIR for medium volume series, Umicore is opening the first high volume factory entirely dedicated to GASIR optics for driving vision enhancement (DVE). This new facility will have a capacity of several tens of thousands of optics per year.
In the low-to-mid IR wavelength range there is a need for high performance, cost effective aspheric optics. Silicon has many advantages including high transmission and a high refractive index, but it can be very difficult to diamond turn. The resulting fabrication errors reduce efficiency and increase scattering and stray light. Wafer-based lithographic techniques can be used to make diffractive and refractive elements in both silicon and germanium. Advantages of diffractive structures such as: thinner elements, highly aspheric and even non-rotationally symmetric phase functions and chromatic compensation make this an attractive technology compared to diamond turning. In addition, wafer based fabrication makes these elements cost-effective in many applications. At Digital Optics Corporation, we have designed and fabricated wafer-based optics for use in the 1.3-14 micron range. In this paper, we will discuss the design, fabrication and evaluation of several product categories including a diffractive germanium beamshaper, a diffractive silicon aspheric lens, and a diffractive silicon spiral lens.
Pupil plane encoding enables extended depth of field and greatly reduced sensitivity to aberrations in an imaging system (field curvature, thermally induced defocus, astigmatism, etc.). The application of pupil plane encoding has potential in thermal imaging where it can enable the use of simple, low-cost, light-weight lens systems. We present numerical and modelling studies of the application of this technique to an uncooled LWIR imaging system, F/1, 75mm focal length, germanium singlet with a detector array size of 240x320 with 50 micron pixel. The initial singlet is corrected from coma and spherical aberration, but its performance across the field of view is greatly limited by astigmatism. The introduction
of an encoding asymmetrical germanium phase mask at the aperture stop of the system, combined with digital image processing, allows the removal of astigmatism and improved imaging performance across the field of view. This improvement is subject to a noise amplification in the digitally restore image. There is as a tradeoff between the maximum correction to astigmatism and reduced signal-to-noise ratio in the recovered image.
The use of lasers as probe sources is very extended in micro and nano technologies. Therefore, the characterization of the beam is critical for the utter development of the measurement. Typically, the beam is projected on the detectors using optical elements and lenses. The alignment procedure is not always very good, and the difficulties increases when infrared radiation is involved. Even with very accurate positioning elements some misalignments are produced. The misalignment is most responsible for the appearance of coma aberration. In the case of a pure Gaussian beam shape they are going to produce a slightly comatic aberrated beam. In this paper we propose a method to characterize the direction and amplitude of this comatic aberration. The method is sensible enough to characterize slightly aberrated beams normally used to deconvolve detector's spatial response. It is based on a statistical analysis of the beam shape in different directions respect to its center. Simulations including the effect of noise are presented too and some applications to micro and nano metrology are exposed.
Infrared antennas are a novel type of detectors that couples electromagnetic radiation into metallic structures and feed it to a rectifying element. As their radio and millimeter counterparts, they can be characterized by parameters explaining their response in a variety of situations. The size of infrared antennas scales with the detected wavelength. Then, specifically designed experimental set-ups
need to be prepared for their characterization. The measurement of the spatial responsivity map of infrared antennas is one of the parameters of interest, but other parameters can be defined to
describe, for example, their directional response, or polarization response. One of the inputs to measure the spatial responsivity map is the spatial distribution of the beam irradiance illuminating
the antenna-coupled detector. The measured quantity is actually a map of the response of the detector when it moves under the beam illumination. This measurement is given as the convolution of the actual responsivity map and the beam irradiance distrbution. The uncertainties, errors, and artifacts incorporated along the measurement procedure are analyzed by using the Principal Component Analysis (PCA). By means of this method is possible to classify different sources of uncertainty. PCA is applied as a metrology tool to characterize the accuracy and repeatability of the experimental set-up. Various examples are given to describe the application of the PCA to the characterization of the deconvolution procedure, and to define the responsivity and the signal-to-noise ratio of the measured results.
Carbon nanotubes (CNT) have a potential to be efficient infrared (IR) detection materials due to their unique electronic properties. The ballistic electronic transport property makes the noise equivalent temperature difference smaller compared to other semi-conducting materials. By overlaying CNT-based mid-IR (3-5μ) detectors on a
long-wave IR (8-15μ) focal plane array, the mid-IR detector causes no filters loss. In order to verify this approach, a single pixel CNT- based infrared photodetector is fabricated by depositing the CNTs on the substrate surface and then aligning them using the atomic force microscopy (AFM)-based nanomanipulation system. Functionality of the single pixel CNT infrared detector is then verified and dark current is analyzed experimentally.
Uncooled infrared focal plane arrays are being developed for a wide range of thermal imaging applications. Firefighting, predictive maintenance, process control and thermography are a few of the industrial applications which could take benefit from uncooled infrared detector. Therefore, to answer these markets, a 35 μm pixel-pitch uncooled IR detector technology has been developed enabling high performance 160 x 120 and 384 x 288 arrays production. Besides a wide-band version from uncooled 320 x 240 / 45 μm array has been also developed in order to address process control and more precisely industrial furnaces control. The ULIS amorphous silicon technology is well adapted to manufacture low cost detector in mass production. After some brief microbolometer technological background, we present the characterization of 35 μm pixel-pitch detector as well as the wide-band 320 x 240 infrared focal plane arrays with a pixel pitch of 45 μm.
In this paper, a readout circuit (ROIC) utilizing a novel noise tolerant edge detection technique for InSb medium wavelength infrared focal plane arrays (MWIR FPAs) is studied. The use of a noise tolerant edge detection algorithm eliminates the need for a pixel-level non-uniformity correction circuit. In addition, the proposed circuit's simple structure allows the processing circuits to be integrated within a shared 2 by 2 pixel area. The proposed method shows better performance for the Gaussian and salt & pepper noise than other conventional approaches. A good edge map is obtained in general InSb MWIR detectors which have 99.5% operability and about 5% non-uniformity of the pixel current. Basic operation of the fabricated noise tolerant edge detection circuit is demonstrated.
We have developed a microbolometer readout integrated circuit (ROIC) that corrects the non-uniformity in analog operation and acts in both normal mode and edge detection mode. A capacitive transimpedance amplifier (CTIA) has been employed as the input circuit of the microbolometer. Generally, when fabricating microbolometer focal plane arrays (FPAs), offset-error and gain-error in the inter-microbolometer are induced by fabrication error. They are shown as fixed pattern noise (FPN) in the infrared image. In the present study, a circuit correcting the offset-error and the gain-error in the normal mode by controlling the bias and the integration capacitance of the CTIA is proposed. This circuit does not require an additional DSP chip, and the non-uniformity is corrected before the analog to digital conversion (ADC). Thus, it can utilize 3-4 bits lower ADC compared to the conventional readout circuit. In the edge detection mode, after correcting the gain-error in two adjacent pixels, edge detection can be realized by subtracting their signal without the DSP. We have designed the suggested circuit to output a 10bit level effective infrared signal using 0.35um 2-poly 3-metal CMOS technology.
Based on the requirements in several applications of object and surface remote identification, and considering the advantages of using multispectral techniques, several systems that allow image acquisition in both specific subbands and single wavelengths have been developed by our group. These systems are based on different techniques. They comprise visible and NIR ranges, with different spectral resolution. Three experimental setups have been developed. The first system is a camera with a filter wheel to choose different spectral bands. The second setup consists of a high-speed camera in which a 1 nm-resolution liquid crystal tunable filter has been assembled. The full system is automatic and allows a fast scan of visible subbands. The third setup uses the same imaging sensor as system #2, but in this case the filter has been substituted by a slit-spectrograph which splits the visible radiation into the different wavelengths that compose the small area observed. The desired wavelength is therefore selected by extracting the appropriate columns of the image acquired from the sensor. The correlation between wavelengths and the CCD array is determined in previous calibration steps. An additional rotatory stage allows the scanning of scenes. Software has been developed to control the systems and make automatic measurements. A new file format specially developed for this project allows the storage of all the images acquired in a single file, which allows a faster ulterior spectral analysis. A bands selection application simplifies the image acquisition depending on the observed scene. The images obtained by the systems will be analyzed in some subsequent stages: qualitative and behavioural study of the elements in the scene, comparison of resolution and operation capabilities of the different configurations and image calibration.
A system was developed to acquire and analyze the rocket plume in the visible and near-infrared bands. The system is designed to be deployed on the field were live rocket firing take place. It is composed of three spectrometers covering the spectrum from 400 nm to 1700 nm, a pair of fast detectors giving the temporal signature at a chosen wavelength band and a pair of CMOS cameras that capture the two-dimensional intensity map of the plume rocket. A detailed description of the instruments that composed the system is described. Results and possible improvement are also presented.
We present a method to evaluate point target detection algorithms. For any particular algorithm, a datacube without a target is evaluated with each pixel being assigned a score; the highest scores belong to those points which are potential false alarms. We then systematically implant a selected target signature into every pixel in the image and evaluate the resulting scores; the lowest scores are those pixels in which the target may be missed. ROC curve analysis can then be made. In this paper, we evaluate a new algorithm which we have developed; we use the evaluation of this algorithm as a paradigm for the efficacy of our algorithm evaluation tool.
Over the last few years, we have developed an algorithm which detects anomalous targets in hyperspectral or multispectral images. The algorithm takes a data (image) cube with a completely unknown background, segments the cube, assigns the largest clusters as background, and determines which pixels are anomalous to the background. In the work to be presented here, we will add two additional modules. First, since our present mission is to detect military targets in a fairly barren rural background, we use the SAVI (or NDVI) metric to detect items which appear to contain chlorophyll. In this way, we can eliminate objects which in retrospect were the right sizes and shapes but were in reality plants. Second, we have developed CFAR methods to achieve a Constant False Alarm Rate while giving us the maximum probability of detecting the targets. Actual data will be analyzed by the algorithm; the ability to both determine if a target is present and where its location is will be shown.
The performance of target detection based on hyperspectral imaging is dependent on the spatial resolution of the imaging sensor. To investigate this dependence, we study a specially prepared scene containing a large number of simple, homogeneous targets in varying sizes. The scene is imaged from the air at a high spatial resolution using the ASI hyperspectral camera. Images corresponding to lower spatial resolutions are synthesized from the recorded data. Targets are detected using spectral anomaly detection based on Gaussian mixture distribution models. The target detection performance is compared between targets of different sizes in images of varying spatial resolution. The results indicate that the spatial resolution of a hyperspectral target detection system should be chosen just high enough to ensure that pure target pixels are present in the image data.
This research addresses the problem of tracking a moving point target from a time sequence of images. The images consist of targets moving at sub-pixel velocity in backgrounds which are influenced by both evolving clutter and noise. The demand for a low false alarm rate on one hand and a high probability of detection on the other makes the tracking a challenging task. We propose the following algorithm in which target detection and tracking is performed by the means of time domain processing. A variance-filter based algorithm is used to detect the presence of targets from the temporal profile of each pixel while suppressing clutter specific influences.