This PDF file contains the front matter associated with SPIE Proceedings Volume 10187 including the Title Page, Copyright information, Table of Contents, Introduction, and Conference Committee listing.

X-ray coherent scatter imaging has the potential to improve the detection of liquid and powder materials of concern in security screening. While x-ray attenuation is dependent on atomic number, coherent scatter is highly dependent on the characteristic angle for the target material, and thus offers an additional discrimination. Conventional coherent scatter analysis requires pixel-by-pixel scanning, and so could be prohibitively slow for security applications. A novel slot scan system has been developed to provide rapid imaging of the coherent scatter at selected angles of interest, simultaneously with the conventional absorption images. Prior experimental results showed promising capability. In this work, Monte Carlo simulations were performed to assess discrimination capability and provide system optimization. Simulation analysis performed using the measured ring profiles for an array of powders and liquids, including water, ethanol and peroxide. For example, simulations yielded a signal-to-background ratio of 1.63±0.08 for a sample consisting of two 10 mm diameter vials, one containing ethanol (signal) and one water (background). This high SBR value is due to the high angular separation of the coherent scatter between the two liquids. The results indicate that the addition of coherent scatter information to single or dual energy attenuation images improves the discrimination of materials of interest.

X-ray inspection systems play a critical role in many non-destructive testing and security applications, with systems typically measuring attenuation during transmission along straight-line paths connecting sources and detectors. Computed tomography (CT) systems can provide higher-quality images than single- or dual-view systems, but the need to measure many projections through the scene increases system complexity and cost. We seek to maximize the image quality of sparse-view (few-view) systems by combining attenuation data with measurements of Compton-scattered photons, that deflect after scattering and arrive at detectors via broken ray paths that provide additional sampling of the scene. The work below presents experimental validation of a singlescatter forward model for Compton-scatter data measured with energy-resolving detectors, and demonstrates a reconstruction algorithm that combines both attenuation and scatter measurements. The results suggest that including Compton-scattered data in the reconstruction process can improve image quality for few-view systems.

X-ray diffraction tomography (XRDT) enables material identification throughout a volumetric object. We perform a simulation-based study to analyze how the image quality of a fan beam coded aperture XRDT system depends on the specifications of the detector used to measure the scatter signal. Going beyond simulation, we design and implement an XRDT prototype scanner that operates in near real-time to perform slice-by-slice imaging of a target object. The scanner is consistent with the requirements of airport checkpoint lanes and can run inline with existing transmission-based X-ray scanners.

Anomaly detection requires a system that can reliably convert measurements of an object into knowledge about that object. Previously, we have shown that an information-theoretic approach to the design and analysis of such systems provides insight into system performance as it pertains to architectural variations in source fluence, view number/angle, spectral resolution, and spatial resolution.¹ However, this work was based on simulated measurements which, in turn, relied on assumptions made in our simulation models and virtual objects. In this work, we describe our experimental testbed capable of making transmission x-ray measurements. The spatial, spectral, and temporal resolution is sufficient to validate aspects of the simulation-based framework, including the forward models, bag packing techniques, and performance analysis. In our experimental CT system, designed baggage is placed on a rotation stage located between a tungsten-anode source and a spectroscopic detector array. The setup is able to measure a full 360° rotation with 18,000 views, each of which defines a 10 ms exposure of 1,536 detector elements, each with 64 spectral channels. Measurements were made of 1,000 bags that comprise 100 clutter instantiations each with 10 different target materials. Moreover, we develop a systematic way to generate bags representative of our desired clutter and target distributions. This gives the dataset a statistical significance valuable in future investigations.

X-ray imaging for security screening is a challenging application that requires simultaneous satisfaction of seemingly incompatible constraints, including low cost, high throughput, and reliable detection of threats. We take a principled computational imaging approach to system design. Mathematical models of the underlying physics and a Huber-class penalty function yield a penalized maximum-likelihood problem. We extend our iterative algorithm for computing linear attenuation coefficients to use multiple energy bins in the SureScan x1000, which has an unconventional, fixed-source geometry. The goal is to maintain the spatial resolution of the single-energy reconstruction while providing information for material characterization used for detection of threats.

Existing approaches to automated security image analysis focus on the detection of particular classes of threat. However, this mode of inspection is ineffectual when dealing with mature classes of threat, for which adversaries have refined effective concealment techniques. Furthermore, these methods may be unable to detect potential threats that have never been seen before. Therefore, in this paper, we investigate an anomaly detection framework, at X-ray image patch-level, based on: (i) image representations, and (ii) the detection of anomalies relative to those representations. We present encouraging preliminary results, using representations learnt using convolutional neural networks, as well as several contributions to a general-purpose anomaly detection algorithm based on decision-tree learning.

X-ray CT based baggage scanners are widely used in security applications. Recently, there has been increased interest in view-limited systems which can improve the scanning throughput while maintaining the threat detection performance. However as very few view angles are acquired in these systems, the image reconstruction problem is challenging. Standard reconstruction algorithms such as the filtered backprojection create strong artifacts when working with view-limited data. In this work, we study the performance of a variety of reconstruction algorithms for both single and multi-energy view-limited systems.

General Electric has designed an innovative x-ray photonic device that concentrates a polychromatic beam of diverging x-rays into a less divergent, parallel, or focused x-ray beam. The device consists of multiple, thin film multilayer stacks. X-rays incident on a given multilayer stack propagate within a high refractive index transmission layer while undergoing multiple total internal reflections from a novel, engineered multilayer containing materials of lower refractive index. Development of this device could lead to order-of-magnitude flux density increases, over a large broadband energy range from below 20 keV to above 300 keV. In this paper, we give an overview of the device and present GE’s progress towards fabricating prototype devices.

X-ray based systems have been used for decades to screen luggage or cargo to detect illicit material. The advent of energy-sensitive photon-counting x-ray detectors mainly based on Cd(Zn)Te semi-conductor technology enables to improve discrimination between materials compared to single or dual energy technology. The presented work is part of the EUROSKY European project to develop a Single European Secure Air-Cargo Space. “Cargo” context implies the presence of relatively heavy objects and with potentially high atomic number. All the study is conducted on simulations with three different detectors: a typical dual energy sandwich detector, a realistic model of the commercial ME100 multi-energy detector marketed by MULTIX, and a ME100 "Cargo": a not yet existing modified multi-energy version of the ME100 more suited to air freight cargo inspection. Firstly, a comparison on simulated measurements shows the performances improvement of the new multi-energy detectors compared to the current dual-energy one. The relative performances are evaluated according to different criteria of separability or contrast-to-noise ratio and the impact of different parameters is studied (influence of channel number, type of materials and tube voltage). Secondly, performances of multi-energy detectors for overlaps processing in a dual-view system is accessed: the case of orthogonal projections has been studied, one giving dimensional values, the other one providing spectral data to assess effective atomic number. A method of overlap correction has been proposed and extended to multi-layer objects case. Therefore, Calibration and processing based on bi-material decomposition have been adapted for this purpose.

X-ray diffraction is known to be an effective technique for illicit materials detection in baggage screening, as it can reveal molecular structural information of any solid substances but also of liquids, aerosols and gels. Some X-ray diffraction systems using 2D pixelated spectrometric detectors, such as CdZnTe detectors, are then able to perform 3D baggage scanning in time compatible with bag throughput constraints of airports. However, X-ray diffraction systems designed for baggage screening generally suffer from poor photon count statistics and bad spatial resolution, because of the tight collimations and the small scattering angle. To improve these factors, techniques of sub-pixelation can be implemented in CdZnTe detectors. Indeed, sub-pixelation enables to open the collimation without angular resolution degradation and also to segment the inspected volume in several sub-volumes, inducing a better spatial resolution in the X-ray beam direction. In this paper, we present some experiments demonstrating the interest of sub-pixelation within CdZnTe detectors for X-ray diffraction imaging systems. In particular, an experimental demonstration is presented with a 2D XRD image of a realistic baggage performed with only one single pixel from our own CdZnTe based imager.

X-ray diffraction (XRD) is well-known for yielding composition and structural information about a material. However, in some applications (such as threat detection in aviation security), the properties of a material are more relevant to the task than is a detailed material characterization. Furthermore, the requirement that one first identify a material before determining its class may be difficult or even impossible for a sufficiently large pool of potentially present materials. We therefore seek to learn relevant composition-structure-property relationships between materials to enable material-identification-free classification. We use an expert-informed, data-driven approach operating on a library of XRD spectra from a broad array of stream of commerce materials. We investigate unsupervised learning techniques in order to learn about naturally emergent groupings, and apply supervised learning techniques to determine how well XRD features can be used to separate user-specified classes in the presence of different types and degrees of signal degradation.

Previously, we investigated the use of Convolutional Neural Networks (CNNs) to detect so-called Small Metallic Threats (SMTs) hidden amongst legitimate goods inside a cargo container. We trained a CNN from scratch on data produced by a Threat Image Projection (TIP) framework that generates images with realistic variation to robustify performance. The system achieved 90% detection of containers that contained a single SMT, while raising 6% false positives on benign containers. The best CNN architecture used the raw high energy image (single-energy) and its logarithm as input channels. Use of the logarithm improved performance, thus echoing studies on human operator performance. However, it is an unexpected result with CNNs. In this work, we (i) investigate methods to exploit material information captured in dual-energy images, and (ii) introduce a new CNN training scheme that generates ‘spot-the-difference’ benign and threat pairs on-the-fly. To the best of our knowledge, this is the first time that CNNs have been applied directly to raw dual-energy X-ray imagery, in any field. To exploit dual-energy, we experiment with adapting several physics-derived approaches to material discrimination from the cargo literature, and introduce three novel variants. We hypothesise that CNNs can implicitly learn about the material characteristics of objects from the raw dual-energy images, and use this to suppress false positives. The best performing method is able to detect 95% of containers containing a single SMT, while raising 0.4% false positives on benign containers. This is a step change improvement in performance over our prior work

Detecting the presence of hazardous materials in suitcases and carry-on luggage is an important problem in aviation security. As the set of threats is expanding, there is a corresponding need to increase the capabilities of explosive detection systems to address these threats. However, there is a lack of principled tools for predicting the performance of alternative designs for detection systems. In this paper, we describe an approach for computing bounds on the achievable classification performance of material discrimination systems based on empirical statistics that estimate the f-divergence of the underlying features. Our approach can be used to examine alternative physical observation modalities and measurement configurations, as well as variations in reconstruction and feature extraction algorithms.