This PDF file contains the front matter associated with SPIE Proceedings Volume 11740, including the Title Page, Copyright information, and Table of Contents.

Time and resource constraints often limit the number of cameras available to establish statistical confidence in determining if a device meets a desired range performance requirement. For thermal cameras, measurements of sampling/resolution, sensitivity and temporal response are combined through the Targeting Task Performance (TTP) metric to predict range. To accommodate a large volume of cameras, we utilized a rotation stage to iterate across the required measurements, with only a single connection instance to the camera. Automation in collection, processing, and device communication reduces opportunities for human error, further improving confidence in the results. To accommodate variations in mounting, cameras were automatically registered to the measurement setup, ensuring accurate analysis and facilitating automatic processing. Additional efficiency was accomplished through processing the measurements in parallel with data collection, reducing the time for full analysis of a single camera from 30 minutes down to 4 minutes. From this work, a statistically relevant sampling of range was accumulated, along with other metrics, to gain insight into manufacturing repeatability, correlated metrics, and datasets for device emulation. In support of the reproducible research effort, many of the analysis scripts used in this work are available for download at [1].

The development and performance verification of a cooled long wave infrared (LWIR) imager optics for a high resolution 10μμm pitch detector is described for the next generation optronics mast systems (OMS). The optical system features a Field of View (FOV) changing re-imager architecture, offering high definition imagery over a 3x magnification range under harsh environmental and built-in conditions, characteristic of submarine periscope applications. Details concerning optical design philosophy and evolution of the system from low (320x256) to high resolution (1024x768) detectors are discussed. The optical system includes a steerable de-scanner plate that enables motion blur compensation in a fast azimuth scan mode of the system for panoramic image acquisition. A conceptual framework simulates the complete imaging path taking into account a combination of relative illumination, distortion and relative boresight error across the FOV's of the system. Systematic limitations of the achievable optical performance due to metrology assisted alignment processes are analyzed with ray-trace modelling. Optical performance metrics of as-built systems from the OMS family are studied from a predictive modelling perspective to qualitatively understand their dominant error modalities. These are used to recommend actions to maximize achievable as-built optical performance for the system under development.

A 1280 x 720 format, 8 μm pixel pitch digital readout integrated circuit (DROIC) for intra-frame high dynamic range (HDR) infrared imaging is presented. Unlike traditional inter-frame HDR imaging where frames with different integration times are temporally combined to obtain an HDR frame, intra-frame HDR imaging is accomplished by spatially interpolating neighboring pixels with different integration times to obtain HDR pixels, thereby achieving the same level of dynamic range improvement without compromising temporal resolution and mostly retaining spatial resolution. In intraframe HDR mode, the infrared imager can achieve a phenomenal >57 dB improvement in dynamic range over normal mode

Authors will report on the optical performance of a newly developed UV display tile made exclusively from COTS components including UVLEDs. Individually addressed, current controlled LEDs with 16 bit gray scale depth make possible a high quality display with frame rate of 1000 Hz, 95% uniformity, and greater than 99% emitter operability. Tile is seamless and can be combined with additional tiles to form a large panoramic display. Although designed for near UV applications, system can be extended to support multispectral application from deep UV through short wave IR.

Low light level (LLL) color vision is becoming more obtainable with improvements in Complementary Metal Oxide Semiconductor Sensor (CMOS) imagers. It is drawing interest in the commercial, military, and first responder communities. Since human perception under LLL conditions is mainly monochrome, what level of color determination is sufficient? For example, is a simple red, yellow, and green street light determination good enough, or is additional color rendering required? Does a speeding car need to be identified as red, or as cherry red? Does a Soldier need to determine if the liquid on a uniform is blood, oil, or water? Additionally, the operator needs to understand that adding specific color capabilities may come at the expense of system performance under LLL conditions. This project seeks to obtain user input, discuss the operating conditions during which personnel would want to have color information, and discuss measurement techniques and methodology standards for color LLL cameras (such as color fidelity, L*a*b* color space, signal-to-noise measurements, luminance uniformity, and color accuracy) that can be utilized by interested stakeholders. We will also discuss the filtering of an incandescent source to create moon-like spectral emission in our laboratory. Preliminary results from color low light camera evaluations at the C5ISR – Night Vision and Electronic Sensors Directorate (NVESD) will be presented.

There are powerful image enhancement methods using, e.g., contrast stretching or retinex to compute enhanced LWIR or MWIR grayvalue images. But some methods have considerable computation times. Additionally, some applications demand object signatures to remain structurally unchanged for further evaluations. Furthermore, since humans can distinguish only far less than 100 grayvalues, but numerous different colors, grayvalue visual- ization of thermal images is considerably suboptimal anyway. So, there are several reasons to use pre-defined, fast color tables. But this pseudo-coloring method is often limited in practice, especially when structures with small thermal differences need to be examined. As a solution, this paper explores different possibilities to reach a maximum benefit for detail uncovering by generating high dynamics in color, saturation, and/or brightness. The challenge is to balance detail uncovering with local as well as global image understandability and to find a long discrete 1D path in 3D color space without visible ambiguities. On the one hand, neighboring colors on this path need to have distances in color space big enough for good detail uncovering. On the other hand, non-neighbor parts of the path need to have even bigger color distances to distinguish high temperature differences. Different color tables are generated and compared. The given results show the reached benefit.

The output voltage of the capacitor (C) and resistor (R) is subject to a low and high pass filter in a RC circuit, respectively. When the input voltage is sinusoidal over time its output voltage is obtained analytically. Its magnitude and phase are determined as a function of the input frequency and the time constant RC. When the input voltage is a Gaussian pulse the RC circuit is no longer a simple low and high pass filter. No analytical solution for the output voltage exists due to the lack of the analytical solution for the integral of the Gaussian function. This makes it cumbersome to predict the output voltage as the ratio of its pulse width to RC. I present an analytical solution in two ways; by using an error function and by approximating the Gaussian function. The latter solution provides the magnitude of the output voltage as a function of the ratio. This is useful in designing a RC circuit for the Gaussian pulse. For example, it can be applied to a pulse detector made of the photoconductor PbSe (Thorlab, PDA20H). An example illustrates such benefit.

A Deep Echo State Neural Network is used to predict total intensity at a detector, standard deviation of intensity over the area of a detector, and center-of-intensity for a deep turbulence example. A short description of the reason for choosing a Deep Echo State Network, as well as a full description of the network optimization and an example using 30 seconds of data is given. Specifically, indications are that this type of network can handle the nonstationary and nonlinear aspects of laser propagation through long distance deep atmospheric turbulence. The network shows a remarkable ability to predict future signals. At this time, more work needs to be done on optimizing the network to achieve even better results.

In this paper we will analyze the effect of vibration on the performance of the YOLO (You Only Look Once) algorithm on object classification. The YOLO algorithm will initially be trained using static imagery of common objects. Image processing techniques will then be used to create video clips with varying levels of blur due to vibration. These videos will then be used to evaluate the performance of the algorithm on object classification. Results of the classification will be presented. An initial summary of the classification results will be discussed in relation to the amount of vibration added to the imagery.

This research investigates the usage of deep neural networks (DNNs) to optimize encoding masks that, when combined with digital imaging systems, can recover high spatial frequency content that would otherwise be filtered out by the OTF. It has been shown that Fourier ptychographic photography (FPP) can recover high resolution imagery by imaging scenes that have been illuminated by a controlled light source. Instead of a controlled light source, this research utilizes light encoding masks that are optimized by a DNN to allow for decreased blur and increased resolution of recovered imagery. A masks generator is optimized in a generative adversarial network (GAN) where generated masks are used to recover imagery through a FPP phase recovery algorithm. The masks and recovery algorithm are optimized according to a loss function comparing the recovered imagery to the pristine, undegraded imagery as dictated by a Wasserstein critic, a DNN model. This method creates masks optimal for recovering high frequency spatial information of specific imaged object types based on the training dataset used.

Optimal longwave infrared (LWIR) scene contrast occurs when reflections from other sources are minimized, leaving only thermal emission. Applying contrast enhancement to LWIR imagery based on pixel values' spatial distribution without regard to underlying temperature and emissivity is a non-physics-based approach. For a physics-based approach, something must be known about the temperature or the objects' emissivity under observation. In remote sensing applications, atmospheric conditions are measured allowing for calculated values for downwelling and path radiance to be obtained. Then, an iterative process can be performed using well-established TES algorithms to determine temperature and emissivity within specific bands. In this paper, we propose a method using a three-band LWIR imaging system with a partial sky view to collect in-scene data to apply contrast enhancement based on spectral differences between bands. Unlike traditional contrast enhancement methods, temperature variations between each band are considered and implemented using relatively inexpensive uncooled microbolometer cameras. We detail the process used for calibrating and determining brightness temperatures with sub-band LWIR filtered cameras. Using absolute sky radiance correlated to MODTRAN6 models, we estimate objects' emissivity profiles in a scene and propose an algorithm for applying contrast enhancement.

Large amounts of labelled imagery are needed to sufficiently train Deep Neural Network (DNN) based classification algorithms. In many cases, collecting an adequate training dataset requires excessive amounts of time and money. The limited data problem is exacerbated when military-relevant imagery requirements are imposed. This often requires imagery collected in the infrared (IR) band, as well as, imagery of military-relevant targets; adding difficulty due to scarcity of sensors, targets, and personnel with the ability to capture the data. To mitigate these types of problems, this study evaluates the effectiveness of synthetic data when aided with small amounts of real data for training DNN based classifier algorithms. This study analyzes the efficacy of the YOLOv3 classifier algorithm at detecting common household objects after training on synthetic data created through an image chipping and insertion method. A set of image chips are created by extracting objects from a green screen background which are then used to generate synthetic training examples by pasting them on a variety of new backgrounds. The impact of background variety and addition of small amounts of real data on trained algorithm performance is analyzed.

The availability of large quantities of labeled data is key to the successful application of artificial intelligence and machine-learning (AIML) algorithms. However, live field data is expensive to collect and the atmospheric conditions that affect image quality are impossible to control. The capability to generate realistic, synthetic training data under operational conditions represents a significant opportunity to operationalize AIML algorithms at a significant cost-savings compared to live field collections. This work seeks to improve upon prior NVESD efforts to simulate the degrading effects of atmospheric turbulence on long-range, ground-to-ground imagery. Specifically, we implement a novel Generative Adversarial Network (GAN) architecture – which we named Noise-GAN – that is capable of detecting and then emulating spatially-varying signals directly from image data. We present results from a proof-of-concept study in which we demonstrate the capability of the Noise-GAN learn the spatial statistics of blur and distortion from turbulence-degraded imagery.

The successful application of machine learning algorithms to ground-to-ground, long-range image applications is dependent upon the availability of a training set of imagery that adequately spans the range of relevant degraded environments. As such, NVESD has developed a turbulence simulation algorithm with the intent of generating realistic, long-range, turbulence–degraded imagery. To properly assess the realism of simulated turbulence– degraded imagery, image comparison metrics must be useful in identifying salient aspects of image degradation. The structural SIMilarity (SSIM) index metric has been developed with the idea that the human visual system is responsive to structural information content. Subsequently, the Multi-Scale SSIM (MS-SSIM) index metric was developed to better handle scale-dependence in image degradation, and the Complex Wavelet SSIM (CW-SSIM) index metric was developed in part to mitigate phase shifts which do not contribute to changes in structural information content. In this study, we assess the extent to which SSIM, MS-SSIM and CW-SSIM are able to quantify salient aspects of degradation in simulated long-range imagery and field data with respect to a pristine reference. Additionally, via the MS-SSIM and CW-SSIM metric approaches, we plan to assess the sensitivities of contrast, structure and luminance in NVESD simulated imagery to perturbations in optical turbulence. We then compare these simulated sensitivities to corresponding field data sensitivities with the intent to inform turbulence simulation development efforts.

The accuracy of modelled performance data for EO/IR sensors is often limited by the accuracy with which image processing functions can be represented in system-level models and simulations. This is particularly so for those cases where complex processing functions are required, such as those found in autonomous ATD/R systems. Furthermore, for sensors mounted on moving platforms, variability in the frame-to-frame image quality can dominate the achieved measures of performance and effectiveness during an engagement. An established technique to address this involves the use of image-based simulations which process dynamically changing imagery using representative image processing functions. However, such an approach requires extensive run-times and a large volume of real or synthetic image data, both of which can be prohibitive. An alternative approach is presented here whereby a limited number of images are processed and then used to generate statistically based performance transfer functions using an appropriate interpolation scheme. These transfer functions are then used to represent the output response of the processing chain when the received imagery is subjected to different levels of degradations such as distortion and blurring. Such transfer functions can then be stored in multidimensional look-up tables which can be rapidly accessed by a system-level Monte-Carlo performance simulation. The ability to represent and extract the performance-related transfer functions is dependent upon the image quality metrics and the accuracy of the corresponding parametric model requires careful consideration of the model validation. An example simulation is presented based on an autonomous ATD/R sensor system mounted on an airborne platform. The importance of validation is demonstrated, and the increased run-time benefits are described. The proposed parametric image modelling approach provides sensor system designers with increased confidence in their design and compliance, and this helps reduces the early design risk.

The Photonics Project is a set of web based calculation tools for educational and analytical use. The tools are primarily Python based notebooks that execute as Web based Apps to the user, not requiring any programming knowledge or installation of any software. There will also be some tools showing full mathematical notation that are MathCad based and require the free MathCad plug in. The calculations primarily follow from equations as presented in the Infrared and Electro-Optical Systems textbook by Ron Driggers et al (second edition). They encompass a suite of Radiometric, Optical, and other photonic functionality. Further efforts are ongoing including an active imaging and photonics devices pages. Like Python itself, the site is open to suggestions and collaboration from users and submission of further tools and functionalities. And totally free of charge to all users.

We demonstrate graded index nanoporous fluoropolymer anti-reflection (AR) coatings for plastic optics that enable <0.5% reflectance from acrylic plastic over the 0.4-2 µm wavelength range for incidence angles up to 40º. Coatings tailored for the visible spectrum yield <0.1 % luminous reflectance, effectively rendering double-side coated plastics invisible under room lighting conditions. Adhesion to most optical plastics, pliability that prevents cracking, outstanding chemical and environmental durability, and compatibility with commercial vacuum coating systems should enable this AR technology to find widespread practical use.

Extended experimental verification of the recently introduced high performance sun blocking filter (SBF) design is reported. The SBF is based on a high optical quality sapphire substrate coated on both sides with multilayers including an induced transparency metal layer, protecting satellite based laser communication terminals (LCTs) telescopes against solar irradiation while simultaneously providing high transmission on the used laser wavelength at 1064.55 nm. Low earth orbit (LEO) satellite based long range LCT offer a potential solution for the highest possible data link rates and availability. Laser transmitter (TX) and receiver (RX) units require a special beam expander telescope optics. The performance of the optics can be deteriorated by the stray light and extreme thermal load conditions caused by direct sun illumination in space. Transmitted wavefront performance is a key requirement for the SBF since the total wavefront budget determines the LCTs link budget. The transmitted wavefront error (TWE) of the SBF is investigated with respect to the coating uniformity. Besides the TWE the contribution of birefringent SBF substrate to the polarization extinction budget has been experimentally verified. Polarization extinction enables polarization coding between TX and RX radiation inside the LCT receiver unit. The interferometric measurements and setups of the transmitted wavefront of the SBF in dependence on substrate temperature are presented.

The temperature dependence of optical window materials remains an important issue for the windows and domes community. Concerning the refractive index, current models are empirically based polynomial fits for the Sellmeier model strength and mode location parameters making it impossible to accurately extrapolate beyond the measured temperature range. Thus, the development of a theoretical model for spectral line shifts as a function of temperature is an important goal. Such a model will allow extrapolation as long as the physical mechanisms don’t change. For vibrational modes, a thermal average of the anharmonically shifted energy levels will be investigated and compared to experimental data. The first anharmonic term can be estimated using the Morse potential based multiphonon absorption model. Experimentally the modes red shift which is consistent with what the proposed model should predict.

Near-net dome shaped ZnS window for the infrared region has been investigated using the Spark Plasma Sintering (SPS) technique. A dome with a large diameter of around 140mm and a height of more than 70mm has been successfully developed. The anti-reflection coated dome has optical transmittance of more than 94% and Knoop hardness above 220. We have found that in addition to the use of SPS technique, the refinement of the commercially used ZnS powder and its purification prior to sintering of the ZnS window was effective in enhancing the sintering time and improving the optical and mechanical properties of the window.

With the advances in the field of high speed and low altitude missiles, thermal shock resistance of IR material domes becomes more critical than ever. In such applications the friction between the dome and the air gives rise to severe aerothermal heating, that can elevate the temperature of the dome well above 250°C in a matter of seconds. Designing the ZnS dome's thickness is a result of a tradeoff between several constraints. Rain Shock and hail impact resistance requirements pull us towards higher thickness. On the other hand, thermal shock resistance favors thinners thicknesses. We present here a recent research of the thermal shock resistance of Rafael made CVD ZnS domes. The method of heating chosen was quenching in hot oil. Ten domes of each type were immersed in mildly stirred oil, heated to consecutively elevated temperatures between 100°C-300°C, and let to cool to room temperature after each step. According to the literature, mildly stirred oil has a heat transfer coefficient in the range of h = 250 ~ 1000 W/m²K, hence for a 4 mm thick ZnS dome, we reach a Biot number between β = 0.06 ~ 0.24 . This is well in the range that is representative of missile dome aerothermal heating, i.e. β = 0.01 ~ 5 . All twenty domes survived the immersions without shattering. This implies that the material properties and process parameters of the Rafael dome production line, exhibits high resistance to thermal shock. Further tests are under way.