<title>Night vision modeling: historical perspective</title>

The Close-In Detector (CID) is the vehicle-mounted multi-sensor anti-tank landmine detection technology for the Army CECOM Night Vision Electronic Sensors Directorate (NVESD) Mine Hunter-Killer (MH/K) Program. The CID includes two down-looking sensor arrays: a 20-antenna ground-penetrating radar (GPR) and a 16-coil metal detector (MD). These arrays span 3-meters in front of a high mobility, multipurpose wheeled vehicle (HMMWV). The CID also includes a roof-mounted, forward looking infrared (FLIR) camera that images a trapezoidal area of the road ahead of the vehicle. Signals from each of the three sensors are processed separately to detect and localize objects of interest. Features of candidate objects are integrated in a processor that uses them to discriminates between anti-tank (AT) mines and clutter and produces a list of suspected mine locations which are passed to the neutralization subsystem of MH/K. This paper reviews the current design and performance of the CID based on field test results on dirt and gravel mine test lanes. Improvements in CID performance for probability of detection, false alarm rate, target positional accuracy and system rate of advance over the past year and a half that meet most of the program goals are described. Sensor performances are compared, and the effectiveness of six different sensor fusion approaches are measured and compared.

Using the latest models from the U.S. Army Night Vision Electronic Sensors Directorate (NVESD), a survey of monochrome and color imaging systems at daylight and low light levels is conducted. Each camera system is evaluated and compared under several different assumptions, such as equivalent field of view with equal and variable f/# , common lens focal length and aperture, with high dynamic range comparisons and over several light levels. The modeling is done by use of the Targeting Task Performance (TTP) metric using the latest version of the Night Vision Integrated Performance Model (NV⁸IPM). The comparison is performed over the V parameter, the main output of the TTP metric. Probability of identification (PID) versus range predictions are a direct non-linear mapping of the V parameter as a function of range. Finally, a comparison between the performance of a Bayer-filtered color camera, the Bayer-filtered color camera with the IR block filter removed, and a monochrome version of the same camera is also conducted.

Range parameters of observation devices can be determined on the basis of numerical simulations (NVTherm) or on the basis of measured characteristics. Those measurements can be conducted in both laboratory and field conditions. It is, however, difficult to carry on reliable field measurements of range parameters because they are strongly depended on atmospheric conditions. Thus the laboratory measurements are more favorable option. Analysis of literature and catalogue specifications reveal, that range parameters are given mainly on the basis of Johnson criteria or TTP model. The Johnson criteria has been used since the 50s and most of catalogue range specifications are determined according to it. There are also NATO standards, which describe the measurement procedures and methodology required to define the detection, recognition and identification ranges for standard NATO targets. For the determination of range parameters the following device characteristics must be known: minimal resolvable temperature for thermal imaging devices and minimal resolvable contrast for VIS devices. The TTP model offers a new approach to the determination of range characteristics of observation devices. It has been developed by U.S. Army’s Night Vision and Electronic Sensors Directorate since the year 2000. It was created because the modified Johnson criteria did not yield reliable results in case of modern systems with digital image processing. In order to determine the range parameters using TTP model, the modulation transfer function MTF, presample MTF function, and 3D noise of a tested system must be known as well as its basic design data as optical magnification and display type. The paper describes the measurement stand, measurement methodology and the procedure for the determination of range parameters. The results for thermal and VIS cameras are also presented, and they are analyzed and compared with the results obtained from current methods, including the measurement uncertainty figures. Some suggestions on the methodology of measurements are also given.

The image formation process of forward looking infrared (FLIR) systems differs from system to system with some systems utilizing gamma correction in the electronics. This is done to compensate for display nonlinearity referred to as the display gamma. Automated minimum resolvable temperature (MRT) testing for an automated support system requires modulation transfer function (MTF) measurement data collected at the video output. To compute the MTF the time averaged edge trace is differentiated digitally to produce the line spread function (LSF) and the FFT. The gamma correction introduced in the video signal for display purposes may introduce errors in the MTF measurement and should be removed for a more accurate result. These nonlinear effects are investigated using simulations and actual measurement data from a FLIR system. The results from the analysis provide one with data for designing improved testing procedures. >

MRTD is considered as the most important parameter of thermal imagers. It is a function of a minimum temperature difference between the bars of the standard 4-bar target and the background required to resolve the thermal image of the bars by an observer versus spatial frequency of the target. Both theoretical models and experimental tests presented in available literature show MRTD as a monotonically increasing bijective mathematical function. This paper presents results of measurements of MRTD that contradict this common opinion. These results show that in rare cases MRTD can be a non-bijective function. Imagers having such MRTD function can fulfill formal requirements on MRTD but will perform poorly at real life conditions. Full Text: PDF References STANAG 4349, Measurement of minimum resolvable thermal difference (MRTD) of thermal cameras, 1995. ASTM standard E 1213-2002, Standard Test Method for Minimum Resolvable Temperature Difference for Thermal Imaging Systems G.C. Holst, Testing and evaluation of infrared imaging systems (JCD Publishing Company 2008). G.C. Holst, The Infrared & Electro-Optical Systems Handbook, Vol.3: Electro-Optical System Design, Analysis and Testing, Ch. 4, 206-207 (SPIE 1993). K. Chrzanowski, Testing Thermal Imagers - Practical Guide (Military University of Technology, Warsaw, Poland 2010). Night Vision Thermal Imaging Systems Performance Model, User?s Manual & Reference Guide, U.S Army Night Vision and Electronic Sensors Directorate, AMSEL-RD-NV-MS-SPMD, 2001. R.G. Driggers, A. Van Hodgkin, R.H. Vollmerhausen, P. O'Shea, "Minimum resolvable temperature difference measurements on undersampled imagers", Proc. SPIE 5076, 179 (2003). CrossRef MERX Website DirectLink Phare Program no PL0103.03.01 PL0108.19.01-L002, Purchase and installation of portable thermo-vision cameras, Phare Program No. PL0103.03.02 purchase of 5 surveillance units for helicopters, 2004.

Target recognition performance for an RCA Shottky barrier staring FLIR with and without target motion is presented together with various minimum resolvable temperature (MRT) measurements obtained using both a 4-bar and a modulation perception requirement, i.e. using the strict requirement to see the 4-bars in the MRT bar pattern and the less restrictive requirement to see some modulation in the displayed image. In addition, both recognition and MRT results for a hand-held thermal viewer having approximately the same limiting spatial frequency as the staring system are presented. The various results are compared and examined to determine if the two dimensional aliasing in staring systems significantly degrades recognition performance relative to conventional scanning systems with similar size detectors and to establish which MRT, vertical 4-bar or modulation, or horizontal 4-bar or modulation is the best predictor of recognition performance when used in conjunction with the Night Vision & Electro-Optics Laboratory (NV&EOL) static recognition model.

McDermott Technology, Inc. (MTI) and JX Crystals have developed an innovative design for a compact, 500-watt, thermophotovoltaic (TPV) power supply using diesel fuel. Under a contract with the Defense Advanced Research Projects Agency (DARPA) and managed by the U.S. Army Communications-Electronics Command (CECOM), this design is being reduced to hardware. Prototypes of the two main subsystems, the power converter assembly (PCA) and the burner/emitter/recuperator (BER), have been designed, fabricated, and tested. The PCA uses low-band-gap gallium antimonide (GaSb) photovoltaic (PV) cells for high efficiency and power density. The prototype PCA will be air cooled for system simplicity and portability. However, initial testing was performed on a water-cooled PCA. The BER uses a thermal vaporizer to produce a stable, high-intensity, low-emissions combustion zone inside an impervious emitter. A thermally integrated recuperator is utilized to boost system efficiency by transferring the unused energy in the exhaust stream to the incoming fuel and combustion air. This paper describes the design, testing and performance of the first-generation PCA and BER along with model predictions used for design and evaluation.

This paper describes the initial steps in the development of a brigade-sized wireless internet system for the Army of the Czech Republic (ACR). In co-operation with the ACR and its Military Technical Institute of Electronics (VTUE), the US Army Communications-Electronics Command (CECOM) and the MITRE Corporation are providing advice on systems engineering and system architecture design for the ACR's Ground Forces Tactical Command and Control System (GF TCCS). The objective system will use a high proportion of commercial off-the-shelf networking software, and government off-the-shelf equipment, including military mobile radios and switches. The key constraints, mandated by VTUE, were as follows: the operating system (Windows NT), the net hardware (radios and switching systems), and NATO compatibility. The tasks on which the CECOM/MITRE team is advising are architectural consultation, systems design and engineering development, database design, user interface consultation, and interoperability analysis.

There is a need for an automated architecture engineering tool to support advanced communication networking efforts. Agencies, such as the US Army's Communications-Electronics Command (CECOM) and Program Executive Office Command, Control, and Communications Systems (PEO-C3S) have the requirement for a suite of applications to provide detailed diagramming, definition and analysis. Many COTS and GOTS products are available to perform some of the tasks required in analyzing, designing, developing, and implementing a network. Many tools which are graphically oriented for use in designing a network, do not provide the capability of automated analysis. While many other network analysis tools are not graphical or user-friendly, they lack the ability to store relevant user data. Some of the information to populate the network architectures exists in inventory databases but the data is not easily accessible and does not contain all the information required.

This paper describes the US Army's efforts under the ATR relational template matching program to develop a multi-sensor forward-looking IR (FLIR) and millimter wave (MMW) radar automatic target recognition (ATR) algorithm. The general problem consists of identifying ground targets at low depression angles of less than 6 degrees and ranges of 500 to 6000 m. However, the algorithm is directly applicable to other target sets, including air targets, and ground targets at high depression angles. Our phase I goal was to provide a proof- of-principle demonstration of this new model-based methodology. We did this on simulated FLIR imagery for a three target class problem. A comprehensive test and evaluation was performed at the Night Vision and Electronic Sensors Directorate. Our goal in phase II was to demonstrate this new algorithm approach on realistic field-collected second-generation FLIR imagery with a larger target set consisting of ten ground vehicles. With phase III, our goal was to demonstrate multisensor fusion on FLIR (first generation) fused with laser radar. This was demonstrated on a four-ground-target-class problem. The limitations to first-generation FLIR data and four targets was due purely to restrictions of the field-collected data. Our goal in phase IV is to demonstrate this methodology for FLIR fused with MMW radar data. Phase IV will be a three-year effort, with the first year concentrating on using the MMW radar data for detection, reduction in false alarms, and the provision of accurate range (i.e., scale) to the FLIR algorithm. The goal of the second and third years will be to fuse the MMW data into not only the detection phase of the algorithm but also the recognition and identification portions of the algorithm. This paper presents the general model-based methodology used for the FLIR and the neural network methodology used to provide the FLIR with the MMW detections and scale. Results are provided for the FLIR-only algorithm for short range and moderate range scenarios given an accurate estimate of target range.

This paper describes the Illowra forward looking infrared (FLIR) performance model, which has been developed by the Defence Science and Technology Organisation (DSTO) and British Aerospace Australia (BAeA). The Illowra model enables calculations of system performance to be carried out for a wide range of FLIR technologies including thermal, photoconductive, and photovoltaic detector based systems used in various staring and scanning detector geometries. Illowra can produce images of target/background scenes, showing spatial and noise effects introduced by a FLIR system. The performance of the human observer, in terms of target detection, recognition, and identification probability, is calculated using the Oracle visual performance model developed by British Aerospace, UK. An outline of the structure and operation of the model is given, followed by a description of how Illowra has been used for various applications within DSTO. The paper concludes with a description of some of the most recent enhancements to Illowra.

Development of an online prediction system for soil organic matter and soil moisture content based on multi-modal fusion

: Parsons Engineering Science, Inc. (Parsons ES) was contracted by the U.S. Army Engineer Waterways Experiment Station (WES) to perform time-domain electromagnetic (TDEM) surveys at two sites. The TDEM surveys were conducted as a part of a study on the background characterization of the response of geophysical sensors used for subsurface unexploded ordnance (UXO) detection. This study was sponsored by the Defense Advanced Research Projects Agency (DARPA) in conjunction with WES, Institute for Defense Analyses (IDA), Walcoff and Associates, and Night Vision Electronic Sensors Directorate (NVESD).

: Because of the multiple efforts across DOD, the Maneuver Support Battle Laboratory proposed and the DFP community funded, establishment of the Virtual Simulation Capability for Deployable Force Protection Analysis (VSCDFP) tool set at Ft. Leonard Wood Missouri. This toolset, a partnership with the Night Vision Electronic Sensors Directorate (NVESD), and the Army Research Laboratory--STTC has integrated multiple simulation tools into a common environment to examine the operational relevance and goodness of multiple sensing modalities, software engines and Common Operation Picture systems in a complex threat environment. This tool set can be used to do comparative analysis on system attributes, and provide insights into Tactics, Techniques and Procedures for DFP lines of effort and other DoD and Army related efforts.

Object detection in thermal images is an important computer vision task and has many applications such as unmanned vehicles, robotics, surveillance, and night vision. Deep learning-based detectors have achieved major progress, which usually need large amount of labelled training data. However, labelled data for object detection in thermal images is scarce and expensive to collect. How to take advantage of the large number labelled visible images and adapt them into thermal image domain is expected to solve. This paper proposes an unsupervised image-generation enhanced adaptation method for object detection in thermal images. To reduce the gap between visible domain and thermal domain, the proposed method manages to generate simulated fake thermal images that are similar to the target images and preserves the annotation information of the visible source domain. The image generation includes a CycleGAN-based image-to-image translation and an intensity inversion transformation. Generated fake thermal images are used as renewed source domain, and then the off-the-shelf domain adaptive faster RCNN is utilized to reduce the gap between the generated intermediate domain and the thermal target domain. Experiments demonstrate the effectiveness and superiority of the proposed method.