This PDF file contains the front matter associated with SPIE Proceedings Volume 11750, including the Title Page, Copyright information, and Table of Contents.

Development over the last two decades of electromagnetic induction (EMI) sensing technology has enabled more accurate identification of unexploded ordinance (UXO) and soil properties. False alarm rates and false negative rates for modern EMI sensors have reduced the cost associated with range cleanup dramatically. However, all EMI UXO sensors continue to have at least two problems associated with safer and more cost-effective cleanup. The first problem centers around requiring a human to push or otherwise operate a sensor on the ground potentially in harm’s way rendering the data acquisition phase of remediation costly and dangerous. The second challenge involves the data interpretation side of EMI sensing. The soil conductivity generates a response that can contaminate EMI data, though in some applications ascertaining soil properties are the goal of the sensor. We developed a lightweight time domain EMI sensor aimed at UXO detection and classification suitable for flying on a UAS platform. This sensor is designed to detect larger targets at up to 2 m from the transmitter and receiver coils. Additionally, we created a time domain sensor suitable for ascertaining the conductivity and magnetic permeability of soils in the top several meters, also suitable for UAS deployment. Both sensors have balanced payloads weighing less than 12 pounds. We present the sensors as well as preliminary calibration and field data demonstrating the efficacy and potential of these sensors to render the process of acquiring UXO data and soil property data less expensive, more efficient and safer.

Detecting and locating underground metallic and non-metallic pipes and utilities remains a pressing problem for the US Department of Energy. Old and deteriorating pipes pose a public safety and environmental hazard but often can be difficult to locate due to poor mapping or broken tracing wires. Many geophysical sensing techniques have been applied to the problem, including: acoustic methods, ground penetrating radar, passive magnetic fields, and low frequency electromagnetic fields – each with its own advantages and pitfalls. This paper investigates a new technique for detecting subsurface pipes: high frequency electromagnetic induction (HFEMI) sensing. Utilizing a frequency range of 10 kHz – 15 MHz, HFEMI has been used successfully in the past for detecting and locating low-conducting subsurface targets such as improvised explosive devices (IED). In this paper, we show HFEMI can be used to induce a linear current in a target pipe which produces a secondary electromagnetic field that can be detected by an above-ground magnetic field receiver. Comparisons between numerical and experimental studies are presented for subsurface elongated conductors. The data is inverted and then validated against ground truth.

Buried improvised explosive devices (IEDs), due to their relative ease of construction, availability, and destructive capacity, remains the main and current (and likely future) asymmetric threat directed at US and coalition forces. The IEDs can be hidden anywhere: in vehicles, on animals, planted in roads or strapped to a person and can be deployed everywhere: in a combat environment or in the middle of a busy city. The adaptability of IEDs to almost any situation makes them difficult to detect, identify and neutralize using standard subsurface sensing technologies, such as low frequency electromagnetic induction (EMI, DC to 100kHz) and ground penetrating radar (GPR, operating above 50MHz). Much research over the past few years has been focused on exploring, developing, and building new systems for buried IED detection. One such technology is the high frequency EMI (HFEMI) sensor developed under an Office Naval Research project. As a part of the project, IEDs detection studies were conducted at a DoD test site using the HFEMI system. The objective of this paper is to illustrate subsurface IED targets detection and identification capabilities using the HFEMI data and models. Namely, first the paper demonstrates the HFEMI data sets collected over IED targets, including intermediate conducting, low-metal content targets, and explosive filled voids; Then, the advanced EMI models and signal processing approaches are adapted to the HFEMI data sets; and finally, applicability of the advanced models are illustrated by postprocessing and inverting HFEMI data sets.

This paper investigates the use of the Hough Transform (HT) for the detection of surface wires that are attached to explosives. Wires, when laid, have a linear structure in which HT can be a promising technique to extract the features and detect the wires. We use a step-frequency ground penetrating radar with a co-pole configuration mounted on an air-borne platform to collect the data for the research study. After beamforming for constructing the data image, projecting onto the surface plane and obtaining the edges, HT is applied to extract features for the detection of wire in the projected image. The features for wire detection include orientation, strength and relative strength that correspond to the tilt angle, the length and the confidence of a wire. Experimental results using the data collected from an indoor facility containing several kinds of wires show the effectiveness of HT for wire detection.

Ultra-wideband (UWB) ground penetrating radar (GPR) is an effective, widely used tool for detection and mapping of buried targets. However, traditional ground penetrating radar systems struggle to resolve and identify congested target configurations and irregularly shaped targets. This is a significant limitation for many municipalities who seek to use GPR to locate and image underground utility pipes. This research investigates the implementation of orbital angular momentum (OAM) control in an UWB GPR, with the goal of addressing these limitations. Control of OAM is a novel technique which leverages an additional degree of freedom offered by spatially structured helical waveforms. This paper examines several free-space and buried target configurations to determine the ability of helical OAM waveforms to improve detectability and distinguishability of buried objects including those with symmetric, asymmetric, and chiral geometries. Microwave OAM can be generated using a uniform circular array (UCA) of antennas with phase delays applied according to azimuth angle. Here, a four-channel network analyzer transceiver is connected to a UCA to enable UWB capability. The characteristic phase delays of OAM waveforms are implemented synthetically via signal processing. The viability demonstrated with the method opens design and analysis degrees of freedom for penetrating radar that may help with discerning challenging targets, such as buried landmines and wires.

Improvised explosives can be, in most cases, prepared from relatively easily accessible matters. The preparation procedure is often simple and generally known, or the information can be looked up on the Internet. The spectrum of phases which can be met concerning these materials is extremely wide and until now no systematic attention has been paid to this fact. In the past years the attention of the Institute of Criminalistics and the University of Pardubice concentrated on preparation of selected groups of these matters, to experimental explosions and to analyses of primary phases and post blast residues using wide spectrum of analytical techniques. Last year the organic analysis was focused on the detection of amines. A simple and sensitive method for the determination of free aliphatic and aromatic amines using derivatives agents as a labelling reagent by high-performance liquid chromatography with fluorescence detection (HPLC-FLD) has been developed. Derivatization conditions including reagent concentration, buffer pH, reaction time and temperature were optimized. A fluorescence derivatization - high-performance liquid chromatography (HPLC) method, which enables the femtomole-level detection of analytes, is a powerful tool for the analysis with high sensitivity and selectivity. Inorganic microanalysis was concentrated on analysis of post-blast residues from experimental explosions of explosive pyrotechnical compositions based on potassium nitrate, sodium nitrate, and barium nitrate – fuel. All the information obtained is stored in a specialized software – database, which will be used by specialized security units. All the activities described above are still in progress.

In this study, we present a new, user friendly, easily extensible and platform independent (cross-platform) GPR signal processing tool named GPRStudio. The tool contains numerous pre-processing and post processing techniques such as background subtraction, various filtering, adjustable gain function, visualization of data, and automatic detection of buried objects. The tool has been developed in Python which is a very popular programming language with its rich and versatile free library alternatives. It has been aimed that the tool can be used on Windows, Linux and MacOS. For this reason, Qt5 is used in graphical user interface design. As an innovative approach to existing GPR signal processing tools, the GPRStudio allows users to write and import their own signal processing algorithms coded in Python. Thus, users can easily observe the effects and results of their own algorithms on GPR data. The project-based structure of the GPRStudio allows the user to work on different collections of GPR data without mixing things up and keep detailed log of every processing step. Users can also import multiple raw GPR data to a project. A raw GPR data can be used in multiple series of processing steps. The tool supports common GPR data types, such as GSSI’s “.dzt”, Sensors and Software’s “.dt1”, MALA’s ”.rd3" and “.rd7” and ASCII “.txt”. Processed data can be exported as “.csv”, “.txt” or “.jpg”.

Running applications and platforms as containerized services is a trending technology for either personal computers and cloud systems. It simplifies various tasks of developers such as building all-in-one, ready-to-use development environments. This paper explains how to easily prepare a stand-alone containerized signal processing environment. The container consists of a Jupyter environment, which is a web-based interactive application that allows the user to write live code and visualize. In this paper we also share a ready-to-use containerized GPR signal processing environment.

Modern supervised machine learning for electro-optical and infrared imagery is based on data-driven learning of features and decision making. State-of-the-art algorithms are largely opaque and questions exist regarding their interpretability and generalizability. For example, what are the learned features, what contexts do they work in, and are the algorithms simply memorizing observations and exploiting unwanted correlations or has it learned an internal representation and causal associations that generalize to new environments? Under the hood, current convolutional neural networks (CNN) are sophisticated curve fitters that are sensitive to sampling (volume and variety). This is problematic as collecting data from real systems is often expensive and time consuming. Furthermore, labeling and quality checking of that data can also be prohibitive. As a result, many are looking to augmentation and simulation to efficiently generate more samples. Herein, we focus on ways to combine augmentation and simulation to improve explosive hazard detection. Specifically, we use the Unreal Engine to produce ray traced simulated data sets of environments and emplacements not captured in real data. We also present a new technique, coined altitude modulated augmentation (AMA), that inserts simulated objects into real world background imagery based on metadata to augment new training data. Thus, the goal of AMA is to increase sampling of observed environments. Preliminary results show that the combination of all techniques is best, followed by augmentation, simulation, then real world data.

Numerous real-world applications require the intelligent combining of disparate information streams from sensors to create a more complete and enhanced observation in support of underlying tasks like classification, regression, or decision making. An often overlooked and underappreciated part of fusion is context. Herein, we focus on two contextual fusion challenges, incomplete (limited knowledge) models and metadata. Examples of metadata available to unmanned aerial systems (UAS) include time of day, platform/sensor position, etc., all of which have a potentially drastic impact on sensor measurements and subsequently our decisions derived from them. Additionally, incomplete models limit machine learning, specifically under-sampling of training data. To address these challenges, we investigate contextually adaptive online Choquet integration. First, we cluster and partition the training metadata. Second, a single machine learning model is trained per partition. Third, a Choquet integral is learned for the combination of these models per partition. Fourth, at test/run time we compute the degree of typicality of a new sample to our known contexts. Fifth, our trained integrals are decomposed into a bag of underlying aggregation operators and a new contextually relevant operator is imputed using a combination of the metadata clustering and observation statistics of the integral variables. This process enables machine learning model selection, ensemble fusion, and metadata outlier detection, with subsequent mitigation strategy identification or decision suppression. The above ideas are demonstrated on explosive hazard detection using surrogate data simulated by the Unreal Engine. In particular, the Unreal Engine is used because it provides us with flexibility to explore the proposed ideas across a range of diverse and controlled experiments. Our preliminary results show improved performance for fusion in different contexts and a sensitivity analysis is performed with respect to metadata degradation.

We show mid-infrared detection of chemical warfare simulants using rapidly tuned and broadband mid-infrared laser spectroscopy suite of chemical species with disparate absorption cross-sections. Sarin gas is one of the most lethal chemical weapons with significant impacts at trace concentrations as low as 64 ppbv within a very short exposure time. In this research, we develop theoretical models to design a mid-infrared (8-11 μm) detection system for high-precision sensing of trace chemical-warfare agents. The models are based on absorption cross-sections and theoretical estimation of chemicals using direct-absorption spectroscopy. Due to the extremely hazardous nature of Sarin, Triethyl Phosphate (TEP), which has a very similar structure to Sarin, was chosen as a proxy chemical. TEP is a standard simulant for organophosphate nerve agents like Sarin. We use a combination of direct absorption spectroscopy and wavelength modulation spectroscopy to resolve and detect line-broadened transitions of TEP. Thus, by analyzing the regression slope of the theoretical absorption cross-section and experimental absorbance, TEP concentration can be estimated in a congested molecular spectrum. The absorption cross-section was modeled using Doppler and Voigt profiles.