This PDF file contains the front matter associated with SPIE Proceedings Volume 8546, including the Title Page, Copyright information, Table of Contents, Introduction, and the Conference Committee listing.

Surface Enhanced Raman Spectroscopy measurements on some common military explosives were performed with a table-top micro-Raman system integrated with a Serstech R785 minispectrometer. Serstech R785 is a miniaturised spectrometer suitable for Raman and NIR measurements. Integration of R785 in our table-top system aims to the realization of a portable SERS detector, able to perform in-situ measurements. SERS Spectra were obtained exciting the substance of interest with a 785 nm diode-laser, while these substances were deposited starting from commercial solutions on commercial SERS substrates, to improve the detection sensitivity. The amount of the sampled substance was determined through the analysis of images of the substrate covered with the residue of explosive. In fact, once the solvent is completely evaporated, the residue of explosive was observed to be uniformly distributed on the substrate surface. Images acquired with a Scanning Electron Microscope provided further details of the deposition process showing that a fraction of the active SERS sites are completely covered with the analyte while other sites appear to be empty; from the analysis of the images the sampled quantity was estimated to be about 200 pg. The main Raman features of each substance were clearly identified, the spectral resolution was sufficiently high to clearly distinguish spectra belonging to different substances.

We report on a variety of BaySpec’s newly developed Raman spectrometers and microscopes combining multiple excitation wavelengths and detection ranges. Among those there are the world’s first dual-wavelength near infrared (NIR) and infrared miniature Raman spectral engines built with Volume Phase Gratings (VPG^TM), and the world’s first three-wavelength (532, 785, and 1064-nm) excitation Raman microscope. Having multiple wavelength excitations in one unit offers extreme flexibility and convenience to identify the best laser wavelength and investigate a great variety of real-world samples. In real-world Raman measurements, fluorescence is the biggest obstacle which significantly reduces the quality of the Raman spectra. We demonstrate many examples spanning from explosives to street drugs to conclude that for those samples, 1064-nm Raman is fluorescence-free and best suited for identification. Other types of miniaturized Raman spectrometers have been realized, enabling handheld, portable, or at-line/ on-line applications for real-world sample measurements, such as threat determination of explosives, chemical and biological materials, quality assurance and contamination control for food safety, and forensics such as evidence gathering, narcotics identification, and anti-counterfeiting.

At the Swedish Defence Research Agency, FOI, Raman spectroscopy is used to detect explosives at stand-off distances. A technique based on imaging Raman spectroscopy has been shown to have the potential to detect trace amounts of explosives at stand-off distances. In this publication we provide limits of detection with the current imaging Raman setup for four different substances, cyclotrimethylenetrinitramine (RDX), 2,4-dinitrotoluene (DNT), 2,4,6-trinitrotoluene (TNT), and sulfur. The limits of detection for DNT and TNT were found to be about 0,5 μg while the lowest limit of detection was achieved for sulfur at 200 ng. The detection limit for RDX is 25,9 μg.

Surface enhanced Raman Scattering spectroscopy is a valuable tool for detecting and identifying chemical threats. One difficulty, however, in utilizing its full capabilities is that the spectrum is dependent upon the chemical orientation, and to a lesser extent, concentration. Spectral peaks can shift and even disappear as the concentration of the chemical present varies. A potential solution to this problem is to model the spectrum as a set of random basis functions, with each basis function depending upon a random unobserved parameter. Relating these parameters to the concentration an expected least squares fitting procedure can be implemented. It is shown through computer simulation and some limited testing that the detection and classification performance can be improved over standard approaches that do not take into account this basis variation. The method proposed, however, is completely general. It is a viable alternative to standard least squares procedures whenever the goal is robustness of the procedure.

Lab-on-a-Bubble (LoB) is a new method for SERS (Surface Enhanced Raman Scattering) assays that combines separationand concentration of the assay results. A direct LoB assay is comprised of gold nanoparticles coupled directly to the ~30 μm diameter buoyant silica bubble. The direct LoB method was evaluated with cyanide and 5,5’-dithiobis(2-nitrobenzoic acid) (DTNB). An indirect assay uses the same ~ 30 μm diameter buoyant silica bubble and a silica coated SERS reporter. Both the bubble and SERS reporter are coated with a coupling agent for the analyte. The assay measures the amount of SERS reporter coupled to the bubble through a sandwich created by the analyte. The couling agent could consist of an immunological coupling agent (antibody) or a nucleic acid coupling agent (single strand DNA). The indirect LoB method was examined with Cholera toxin (CT) and antibodies against the β subunit. An LOD of ~ 170 pptrillion was measured for cyanide and a limit of detection of 1100 ng was found for CT.

Instrumentation for the assay and a novel technique of dynamic SERS (DSERS) will also be discussed. The instrument is a small hand-held Raman device called the CBEx (Chemical Biological Explosive) with a novel raster system to detect heterogeneous or light sensitive materials. DSERS is a mathematical algorithm which eliminates background interference in SERS measurements with colloidal nanoparticles.

Spectrally tunable narrow-linewidth mid-infrared sources are used in a variety of spectrometric optical systems for detection, identification, and/or quantification of chemical species. However, in the pulsed regime they often display a varying spectrum in time, either from shot-to-shot or during the pulse itself, with consequences on the measurement accuracy, resolution, and repeatability. This is, for instance, the case of pulsed quantum cascade lasers (QCL), mainly because of strong transient thermal effects in the optical waveguide. Unfortunately, little information has been published on this subject because mid-infrared time-resolved spectrometers are extremely scarce. In this paper, we explain how this can be circumvented by using time-gated frequency upconversion in a nonlinear crystal. We apply this principle to characterize a pulsed external cavity QCL (EC-QCL) at 7.8 μm, using AgGaS₂ as the nonlinear crystal and a Q-switched Nd:YAG laser as the pump source. The upconverted near infrared spectrum is conveniently analyzed with a high resolution lambdameter and an optical spectrum analyzer. We evidence frequency chirp at an average rate of -50 MHz/ns and mode hops spanning 15 GHz for the EC-QCL. These results are compared to published data.

High Power Laser Diodes (HPLD) are increasingly used in different fields of applications such as Industry, Medicine and Defense. Our significant improvements of performances (especially in power and efficiency) and a reproducible manufacturing process have led to reliable, highly robust components. For defense and security applications these devices are used predominantly for pumping of solid state lasers (ranging, designation, countermeasures, and sensors). Due to the drastically falling price per watt they are more and more replacing flash lamps as pump sources. By collimating the laser beam even with a bar to bar pitch of only 400μm. cutting edge brightness of our stacks.is achieved Due the extremely high brightness and high power density these stacks are an enabling technology for the development of compact highly efficient portable solid state lasers for applications as telemeters and designators on small platforms such as small UAVs and handheld devices. In combination with beam homogenizing optics their compact size and high efficiency makes these devices perfectly suited as illuminators for portable active imaging systems. For gated active imaging systems a very short pulse at high PRF operation is required. For this application we have developed a diode driver board with an efficiency several times higher than that of a standard driver. As a consequence this laser source has very low power consumption and low waste heat dissipation. In combination with its compact size and the integrated beam homogenizing optics it is therefore ideally suited for use in portable gated active imaging systems. The kWatt peak power enables a range of several hundred meters. The devices described in this paper mostly operate at wavelength between 800 nm and 980nm. Results from diodes operating between 1300 nm and 1550 nm are presented as well.

An important part of criminalistic forensics is the analysis of toolmarks. Such toolmarks often consist of plenty of single striations, scratches and dents which can allow for conclusions in regards to the sequence of events or used tools. To receive qualified results with an automated analysis and contactless acquisition of such toolmarks, a detailed digital representation of these and their orientation as well as placing to each other is required. For marks of firearms and tools the desired result of an analysis is a conclusion whether or not a mark has been generated by a tool under suspicion. For toolmark analysis on locking cylinders, the aim is not an identification of the used tool but rather an identification of the opening method. The challenge of such an identification is that a one-to-one comparison of two images is not sufficient - although two marked objects look completely different in regards to the specific location and shape of found marks they still can represent a sample for the identical opening method. This paper provides the first approach for modelling toolmarks on lock pins and takes into consideration the different requirements necessary to generate a detailed and interpretable digital representation of these traces. These requirements are ’detail’, i.e. adequate features which allow for a suitable representation and interpretation of single marks, ’meta detail’, i.e. adequate representation of the context and connection between all marks and ’distinctiveness’, i.e. the possibility to reliably distinguish different sample types by the according model. The model is evaluated with a set of 15 physical samples (resulting in 675 digital scans) of lock pins from cylinders opened with different opening methods, contactlessly scanned with a confocal laser microscope. The presented results suggest a high suitability for the aspired purpose of opening method determination.

Fingerprints and microtraces play an important role as evidence within the field of criminalistics. Their conservative acquisition processes, are established, but are altering and impurifying the traces often. In case of microtraces even the integrity of the trace complex is affected. Using contactless methods, the acquisition process becomes non-invasiv and repeatable, but might be distorting on the other hand, when non-planar substrates are in use. Detecting and dealing with distortion in contactless aquired scans of non-planar surfaces is a novel field of research. Nowadays highly distorted fingerprints can only be used, if the substrate can be manually distorted by destroying or deforming it. In this paper we suggest methods for detection and equalization of distortion for use in combination of types of traces. Therefore we define different types of distortion in fingerprints and microtraces. A standardization of types is necessary to develop different solution for equalization. For usage within the field of forensics, each method is evaluated via proper error rates and adaptively used to acquire fingerprints and microtraces. Using our techniques, we are able to detect distortion and equalize fingerprints to support the investigators work. In case of microtraces the presented methods can even be used to equalize mircotraces themselves for better determination of their scale and topology. For all scans the confocal 3D laser microscope "Keyence VK-X110" is used to gather color-, intensity- and topography information in 22 different measurement conditions within 6 different samples consisting of a total of 880 scans. Despite our achievements in the field of distortion detection and equalization there are still challenges, like the non-isometric projection, that need to be focused on. Also, the presented equalization methods may not completely remove any kind of distortion, such as added by deformation. Therefore we suggest and discuss future work for improving the distortion detection process by adding classification of sources of distortion to assure that the correct equalization method is used.

The objective of this research is to develop a new robust fingerprint identification technology based upon forming surface-subsurface (under skin) ultrasonic 3D images of the finger pads. The presented work aims to create specialized ultrasonic scanning methods for biometric purposes. Preliminary research has demonstrated the applicability of acoustic microscopy for fingerprint reading. The additional information from internal skin layers and dermis structures contained in the scan can essentially improve confidence in the identification. Advantages of this system include high resolution and quick scanning time. Operating in pulse-echo mode provides spatial resolution up to 0.05 mm. Technology advantages of the proposed technology are the following: • Full-range scanning of the fingerprint area "nail to nail” (2.5 x 2.5 cm) can be done in less than 5 sec with a resolution of up to 1000 dpi. • Collection of information about the in-depth structure of the fingerprint realized by the set of spherically focused 50 MHz acoustic lens provide the resolution ~ 0.05 mm or better • In addition to fingerprints, this technology can identify sweat porous at the surface and under the skin • No sensitivity to the contamination of the finger's surface • Detection of blood velocity using Doppler effect can be implemented to distinguish living specimens • Utilization as polygraph device • Simple connectivity to fingerprint databases obtained with other techniques • The digitally interpolated images can then be enhanced allowing for greater resolution • Method can be applied to fingernails and underlying tissues, providing more information • A laboratory prototype of the biometrics system based on these described principles was designed, built and tested. It is the first step toward a practical implementation of this technique.

It has been shown that both optical and acoustic signals can be transmitted through clothing. This means that, in principle, both types of signal could be used for personal screening. In this paper, a description of how each of these signals interacts with clothing, and the information that can be obtained, will be presented. The wavelengths are very different - by three orders of magnitude. In fact, the wavelength of ultrasound is of a similar order of magnitude to that used by THz systems (≤ 1 mm), whereas the optical signals that could be used are typically in the micron or so range. This means that their use for screening relies on totally different mechanisms. In this paper, results will be shown for both types of measurements, where the location of hidden objects can be demonstrated. The imaging resolution that can be obtained will be described, together with a discussion of how both types of energy could be used for improved identification of objects hidden beneath layers of clothing.

Bottled liquids are limited to be brought in the airplane, because liquid explosives have been used in some terrorist attaches recently. A bottled liquid scanner is expected to be developed. Liquid scanner using near infrared technologies is being developed by us. Many spectrum of liquids have been collected and analyzed by chemometorics in order to separate safe beverage to explosive and dangerous liquids. This bottled liquid scanner had feasibility tests in some international airport in Japan and obtained good review from security people in the airport.

The feasibility of stand-off detection of liquid surface contamination by means of hyperspectral imaging has been shown in previous studies. Our previous studies focused on measurements using an active set-up in which the sample is illuminated by an infrared radiation source. In this work, passive measurements of spilled liquids are presented that were performed under the open sky and without illumination by an active source. Because the spectra depend on the liquid (composition and layer thickness), the optical properties of the background material, and on the radiation of the surrounding hemisphere that is incident on the sample, a comprehensive radiative transfer model has been developed and applied to calculate spectra of the specific liquid-substrate combination. The comparison of calculated and measured spectra allows the visualization of contaminated areas, the detection, and the identification of the liquids. The agreement between measured and modelled spectra is excellent. This shows that the automatic identification of liquid surface contaminants is possible using a passive setup and that the proposed model is applicable for the approximation of the measured spectra.

We demonstrate the improvement of the quality of the image captured by TS4 - the commercially available THz passive camera manufactured by ThruVision Systems Ltd. The measurements range of this device reaches 10 meters. Our approach is based on application of novel spatial filters and algorithms, developed by us for computer processing of passive THz images produced by the various THz cameras. In our opinion, the most important result of this paper consists in a demonstration of the possibility of using a passive THz camera to observe a difference in temperature on the human skin if this difference is caused by different temperatures in the inside of the body. Such possibility was proposed by Vyacheslav Trofimov on the Conference in Baltimore (April 2012) as well as and in [20]. We discuss two physical experiments, in which a person drinks hot and cold water. After computer processing of images captured by passive THz camera TS4 we may see the pronounced temperature trace on the human body. We illustrate this phenomenon by a series of images captured by passive THz camera in real time. As we believe, these experiments allow us wide applications of passive THz cameras for the detection of objects concealed in the inside of the human body because the difference in temperature that will be reflected on the human skin. Modern passive THz cameras have not enough resolution in temperature to see this difference. However, computer processing allows us to enhance it for this application. Using computer processing one may enhance the image quality and delete noise on the images. In some cases, it is possible to achieve full de-noising of the image.

Currently, there exists a capability gap for the remote detection and identification of threat chemicals. We report here on the development of an Active Coherent Laser Spectrometer (ACLaS) operating in the thermal infrared and capable of multi-species stand-off detection of chemicals at sub ppm.m levels. A bench top prototype of the instrument has been developed using distributed feedback mid-infrared quantum cascade lasers as spectroscopic sources. The instrument provides active eye-safe illumination of a topographic target and subsequent spectroscopic analysis through optical heterodyne detection of the diffuse backscattered field. Chemical selectivity is provided by the combination of the narrow laser spectral bandwidth (typically < 2 MHz) and frequency tunability that allows the recording of the full absorption spectrum of any species within the instrument line of sight. Stand-off detection at distances up to 12 m has been demonstrated on light molecules such as H₂O, CH₄ and N₂O. A physical model of the stand-off detection scenario including ro-vibrational molecular absorption parameters was used in conjunction with a fitting algorithm to retrieve quantitative mixing ratio information on multiple absorbers.

MR-i is a dual band Hyperspectral Imaging Spectro-radiometer. This field instrument generates spectral datacubes in the MWIR and LWIR. MR-i is modular and can be configured in different ways. One of its configurations is optimized for the standoff measurements of gases in differential mode. In this mode, the instrument is equipped with a dual-input telescope to perform optical background subtraction. The resulting signal is the differential between the spectral radiance entering each input port. With that method, the signal from the background is automatically removed from the signal of the target of interest. The spectral range of this configuration extends in the VLWIR (cut-off near 14 μm) to take full advantage of the LW atmospheric window.

Detection and identification of Toxic industrial chemicals (TICs) represent a major challenge to protect and sustain first responder and public security. In this context, passive Hyperspectral Imaging (HSI) is a promising technology for the standoff detection and identification of chemical vapors emanating from a distant location. To investigate this method, the Department of National Defense and Public Safety Canada have mandated Defense Research and Development Canada (DRDC) – Valcartier to develop and test Very Long Wave Infrared (VLWIR) HSI sensors for standoff detection. The initial effort was focused to address the standoff detection and identification of toxic industrial chemicals (TICs), surrogates and precursors. Sensors such as the Improved Compact ATmospheric Sounding Interferometer (iCATSI) and the Multi-option Differential Detection and Imaging Fourier Spectrometer (MoDDIFS) were developed for this application. This paper presents the sensor developments and preliminary results of standoff detection and identification of TICs and precursors. The iCATSI and MoDDIFS sensors are based on the optical differential Fourier-transform infrared (FTIR) radiometric technology and are able to detect, spectrally resolve and identify small leak at ranges in excess of 1 km. Results from a series of trials in asymmetric threat type scenarios are reported. These results serve to establish the potential of passive standoff HSI detection of TICs, precursors and surrogates.

Active hyperspectral imaging is a valuable tool in a wide range of applications. One such area is the detection and identification of chemicals, especially toxic chemical warfare agents, through analysis of the resulting absorption spectrum. This work presents a selection of results from a prototype midwave infrared (MWIR) hyperspectral imaging instrument that has successfully been used for compound detection at a range of standoff distances. Active hyperspectral imaging utilises a broadly tunable laser source to illuminate the scene with light at a range of wavelengths. While there are a number of illumination methods, the chosen configuration illuminates the scene by raster scanning the laser beam using a pair of galvanometric mirrors. The resulting backscattered light from the scene is collected by the same mirrors and focussed onto a suitable single-point detector, where the image is constructed pixel by pixel. The imaging instrument that was developed in this work is based around an IR optical parametric oscillator (OPO) source with broad tunability, operating in the 2.6 to 3.7 μm (MWIR) and 1.5 to 1.8 μm (shortwave IR, SWIR) spectral regions. The MWIR beam was primarily used as it addressed the fundamental absorption features of the target compounds compared to the overtone and combination bands in the SWIR region, which can be less intense by more than an order of magnitude. We show that a prototype NCI instrument was able to locate hydrocarbon materials at distances up to 15 metres.

A real time processor package called IT4 is described, which has been developed to improve the quality of Imaging Through the 4 artefacts of sub-optimal surveillance (jitter, shimmer, contrast and noise). The package is able to operate on a wide range of video standards and can process imagery from the camera sensor directly or from previously recorded (movie file) sequences. The platform is currently a PC or a laptop but no other requirements such as special optics or camera hardware are necessary. IT4 removes both translational and rotational camera movement using a de-jitter algorithm and the image frames are analyzed to discriminate atmospheric shimmer from real image motion. The frames are then de-shimmered but any scene motion is accurately rendered. Sharpening and contrast options are available to the user to further improve visual quality and the contrast can be enhanced locally if so desired. IT4 improves fixed pattern noise in the image at all light levels and also reduces random noise, which is particularly useful during low light level image capture leading to better processing of both static and dynamic scenes.

Real-time super-resolution within surveillance video streams is a powerful tool for security and crime prevention allowing for example, events, faces or objects such number-plates and luggage to be more accurately identified on the fly and from a distance. However, many of the state of the art approaches to super-resolution are computationally too expensive to be suitable for real-time applications within a surveillance context. We consider one particular contemporary method based on sparse coding,¹ and show how, by relaxing some model constraints, it can be sped up significantly compared to the reference implementation, and thus approach real-time performance with visually indistinct reduction in fidelity. The final computation is three orders of magnitude faster than the reference implementation. The quality of the output is maintained: PSNR of the super-resolved images compared to ground truth is not significantly different to the reference implementation, while maintaining a noticeable improvement over baseline bicubic-interpolation approach.

Stand-off base and force protection surveillance measures primarily rely on electro-optic and thermal imaging technology. Atmospheric turbulence causes blur, distortion and intensity fluctuations that can severely degrade the image quality of these systems. This work explores the effects of turbulence image degradation on the performance of automatic facial recognition software and also looks at the potential benefit of turbulence mitigation algorithms. The goal of this work is to understand the feasibility of long-range facial recognition in degraded imaging conditions. In order to create a large enough database to match against, simulated imagery of different ranges and turbulence conditions were created using a horizontal view turbulence simulator and a subset of the Facial Recognition Technology (FERET) database. The simulated turbulence degraded imagery was then processed with facial recognition software and the results are compared against those from the pristine image set. Finally, the performance of the facial recognition software with turbulence mitigated imagery is also presented.

This paper proposes a visual scene busyness indicator obtained from the properties of a full spatial segmentation of static images. A fast and effective region merging scheme is applied for this purpose. It uses a semi-greedy merging criterion and an adaptive threshold to control segmentation resolution. The core of the framework is a hierarchical parallel merging model and region reduction techniques. The segmentation procedure consists of the following phases: 1. algorithmic region merging, and 2. region reduction, which includes small segment reduction and enclosed region absorption. Quantitative analyses on standard benchmark data have shown the procedure to compare favourably to other segmentation methods. Qualitative assessment of the segmentation results indicate approximate semantic correlations between segmented regions and real world objects. This characteristic is used as a basis for quantifying scene busyness in terms of properties of the segmentation map and the segmentation process that generates it. A visual busyness indicator based on full colour segmentation is evaluated against conventional measures.

Performing persistent surveillance of large populations of targets is increasingly important in both the defence and security domains. In response to this, Wide Area Motion Imagery (WAMI) sensors with Wide FoVs are growing in popularity. Such WAMI sensors simultaneously provide high spatial and temporal resolutions, giving extreme pixel counts over large geographical areas. The ensuing data rates are such that either very bandwidth data links are required (e.g. for human interpretation) or close-to-sensor automation is required to down-select salient information. For the latter case, we use an iterative quad-tree optical-flow algorithm to efficiently estimate the parameters of a perspective deformation of the background. We then use a robust estimator to simultaneously detect foreground pixels and infer the parameters of each background pixel in the current image. The resulting detections are referenced to the coordinates of the first frame and passed to a multi-target tracker. The multi-target tracker uses a Kalman filter per target and a Global Nearest Neighbour approach to multi-target data association, thereby including statistical models for missed detections and false alarms. We use spatial data structures to ensure that the tracker can scale to analysing thousands of targets. We demonstrate that real-time processing (on modest hardware) is feasible on an unclassified WAMI infra-red dataset consisting of 4096 by 4096 pixels at 1Hz simulating data taken from a Wide FoV sensor on a UAV. With low latency and despite intermittent obscuration and false alarms, we demonstrate persistent tracking of all but one (low-contrast) vehicular target, with no false tracks.

A user-assisted multi-camera tracking system employing several key novel methodologies has previously been shown to be highly effective in assisting human users in tracking targets of interest through industry-standard i-LIDS multi-camera benchmark data.¹ A prototype system was developed in order to test and evaluate the effectiveness of this approach. In this paper, we develop this system further in order to improve tracking accuracy and further facilitate scalability to arbitrary numbers of camera views across much larger spatial areas and different locations. Specifically, we describe the following three areas of improvement: (1) dynamic learning mechanisms apply user feedback in adapting internal models to improve performance over time; (2) modular design and hardware acceleration techniques are explored with a view to real-time performance, extensive configurability to leverage available hardware and scalability to larger datasets; and (3) re-design of the user interface for deployment as a secure asynchronous remote web-based service. We conduct an extensive evaluation of the system in terms of: (1) tracking performance; and (2) the speed of the system in computation and in usage over a network. We use a newly collected real-world dataset significantly more challenging than i-LIDS, which comprises six cameras covering two London Underground stations. We show that: (1) dynamic learning is effective; (2) the user-assisted paradigm retains its effectiveness with this significantly more challenging dataset; (3) large-scale deployment and real-time computation is feasible due to linear scalability; (4) context-aware user search strategies and external non-visual information can aid search convergence; and (5) storage and querying of meta-data is a bottleneck to be overcome.

The Image Library for Intelligent Detection Systems (i-LIDS) provides benchmark surveillance datasets for analytics systems. This paper proposes a methodology to investigate the effect of compression and frame-rate reduction, and to recommend an appropriate suite of degraded datasets for public release. The library consists of six scenarios, including Sterile Zone (SZ) and Parked Vehicle (PV), which are investigated using two different compression algorithms (H.264 and JPEG) and a number of detection systems. PV has higher spatio-temporal complexity than the SZ. Compression performance is dependent on scene content hence PV will require larger bit-streams in comparison with SZ, for any given distortion rate. The study includes both industry standard algorithms (for transmission) and CCTV recorders (for storage). CCTV recorders generally use proprietary formats, which may significantly affect the visual information. Encoding standards such as H.264 and JPEG use the Discrete Cosine Transform (DCT) technique, which introduces blocking artefacts. The H.264 compression algorithm follows a hybrid predictive coding approach to achieve high compression gains, exploiting both spatial and temporal redundancy. The highly predictive approach of H.264 may introduce more artefacts resulting in a greater effect on the performance of analytics systems than JPEG. The paper describes the two main components of the proposed methodology to measure the effect of degradation on analytics performance. Firstly, the standard tests, using the ‘f-measure’ to evaluate the performance on a range of degraded video sets. Secondly, the characterisation of the datasets, using quantification of scene features, defined using image processing techniques. This characterization permits an analysis of the points of failure introduced by the video degradation.

In this paper we present the multi-sensor registration and fusion algorithms that were developed for a force protection research project in order to detect threats against military patrol vehicles. The fusion is performed at object level, using a hierarchical evidence aggregation approach. It first uses expert domain knowledge about the features used to characterize the detected threats, that is implemented in the form of a fuzzy expert system. The next level consists in fusing intra-sensor and inter-sensor information. Here an ordered weighted averaging operator is used. The object level fusion between candidate threats that are detected asynchronously on a moving vehicle by sensors with different imaging geometries, requires an accurate sensor to world coordinate transformation. This image registration will also be discussed in this paper.

A recent trend in law enforcement has been the use of Forensic lip-readers. Criminal activities are often recorded on CCTV or other video gathering systems. Knowledge of what suspects are saying enriches the evidence gathered but lip-readers, by their own admission, are fallible so, based on long term studies of automated lip-reading, we are investigating the possibilities and limitations of applying this technique under realistic conditions. We have adopted a step-by-step approach and are developing a capability when prior video information is available for the suspect of interest. We use the terminology video-to-text (V2T) for this technique by analogy with speech-to-text (S2T) which also has applications in security and law-enforcement.

Active hyperspectral imaging is a valuable tool in a wide range of applications. A developing market is the detection and identification of energetic compounds through analysis of the resulting absorption spectrum. This work presents a selection of results from a prototype mid-infrared (MWIR) hyperspectral imaging instrument that has successfully been used for compound detection at a range of standoff distances. Active hyperspectral imaging utilises a broadly tunable laser source to illuminate the scene with light over a range of wavelengths. While there are a number of illumination methods, this work illuminates the scene by raster scanning the laser beam using a pair of galvanometric mirrors. The resulting backscattered light from the scene is collected by the same mirrors and directed and focussed onto a suitable single-point detector, where the image is constructed pixel by pixel. The imaging instrument that was developed in this work is based around a MWIR optical parametric oscillator (OPO) source with broad tunability, operating at 2.6 μm to 3.7 μm. Due to material handling procedures associated with explosive compounds, experimental work was undertaken initially using simulant compounds. A second set of compounds that was tested alongside the simulant compounds is a range of confusion compounds. By having the broad wavelength tunability of the OPO, extended absorption spectra of the compounds could be obtained to aid in compound identification. The prototype imager instrument has successfully been used to record the absorption spectra for a range of compounds from the simulant and confusion sets and current work is now investigating actual explosive compounds. The authors see a very promising outlook for the MWIR hyperspectral imager. From an applications point of view this format of imaging instrument could be used for a range of standoff, improvised explosive device (IED) detection applications and potential incident scene forensic investigation.

At the Diagnostic and Metrology Laboratory (ENEA) is under development a stand-off apparatus for trace detection of explosive compounds. The system is based on the Raman technique due to the high discrimination capabilities. All the optoelectronics components of the apparatus have been carefully designed as well as their optical matching. The main goal will be to detect low trace components within the respect of the eye safe regulation.

The most desirable configuration for detection of toxic chemicals utilises the maximum distance between detector and hazard. This approach minimises the contamination of equipment or personnel. Where the target chemical is an involatile liquid, indirect detection of the liquid contamination is made difficult by inherently low vapour pressure. In this instance, direct detection of the chemical hazard is the best approach. Recent technology developments have allowed spectroscopic systems to provide multiple options for the stand-off detection of involatile chemical warfare agents (CWAs). Two different stand-off spectroscopic systems, based upon IR absorption and Raman spectroscopic techniques are described here. The Negative Contrast Imager (NCI) is based upon an optical parametric oscillator (OPO) source comprising a Q-switched intracavity MgO:PPLN crystal. This crystal has a fanned grating design and wavelength tuning is achieved by translating the PPLN crystal within the 1064 nm pump beam. This approach enables the production of shortwave and midwave IR radiation (1.5 – 1.8 μm and 2.6 – 3.8 μm, respectively), which is scanned across the scene of interest. Target materials that have an absorption feature commensurate with the wavelength of incoming radiation reduce the intensity of returned signal, resulting in dark pixels in the acquired image. This method enables location and classification of the target material. Stand-off Raman spectroscopy allows target chemicals to be identified at range through comparison of the acquired signature relative to a spectral database. In this work, we used a Raman system based upon a 1047 nm Nd:YLF laser source and a proprietary InGaAsP camera system. Utilisation of a longer excitation wavelength than most conventional stand-off detection systems (e.g. 532 or 785 nm) enables reduction of fluorescence from both the surface and the deposited chemicals, thereby revealing the Raman spectrum. NCI and Raman spectroscopy are able to detect CWAs on surfaces at distances of 2 – 10 metres and have potential to detect over longer ranges. We report the successful identification of at least 60 μl of nitrogen mustard at a distance of a 2 m and 10 m using NCI and Raman spectroscopy.

TATP is a very easy to synthesize [9], sensitive, high explosive [10] and high volatile explosive [1, 3, 7] with great absorption in the IR Spectra [4, 5, 6]. In this project we detect TATP gas traces with open path FTIR – techniques. The first project phase was to construct and build a heatable multi-reflection cell with adjustable optical path length and a heatable intake to evaporate solid TATP samples. In this cell reference TATP - spectra were collected under controlled conditions with a Bruker FTIR system (Typ OPAG 33). The next step was to find out how the TATP gas will be diluted in the ambient air and validate some physical properties which are described inconsistently in literature e.g. evaporation rates. We constructed a special double - T shaped chamber with stabile air conditions. In this chamber the dispersion kinetics of the TATP vapour could be tested. It turned out that the TATP vapours has the tendency to drop down. Therefore the highest TATP - concentrations were measured below the TATP sample. During the investigation for this study it turned out, that some materials scrub the TATP- vapour out of the air, e.g. Metals, fabric, leather. In the second phase of the project successful open path FTIR- measurements were taken in ambient air and will be continued with different system configurations of the OPAG 33 to lower the detection limits. Also successful measurements were taken in indoor ambient air with a Hyper spectral camera (passive FTIR with array sensor) to detect TATP in solid and gaseous phase. This technique allows detecting TATP and identifying the TATP source. The poster shows some selected results of the continued research.

Fourier Transform infrared spectroscopy (FTIR) is regularly used in forensic analysis, however the application of high resolution Fourier Transform infrared spectroscopy for the detection of explosive materials and explosive precursors has not been fully explored. This project aimed to develop systematically a protocol for the analysis of explosives and precursors using Fourier Transform infrared spectroscopy and basic data analysis to enable the further development of a quantum cascade laser (QCL) based airport detection system. This paper details the development of the protocol and results of the initial analysis of compounds of interest.