Understanding far and close proximity human-human interaction observed from a distance is a necessary step towards automated suspicious or antisocial behaviour detection. Most previous work on human-human interaction has made the implicit assumption that interactions occur only at immediate spatial and temporal proximity between the subjects concerned. We propose a more realistic application of human-human interaction detection from surveillance data where the subjects of interest tend to be represented by few pixels relative to the rest of the scene. The subjects are represented by relatively few pixels since surveillance cameras are usually placed to maximise area coverage therefore there is a significant distance between the camera and the physical scene. This in itself is not so much of a disadvantage when we consider that interactions among subjects can occur between quite large distances in space. Our technique uses a spatial and temporal saliency measure to extract and select features using modifications to Kadir and Brady's scale saliency and Hung and Gong's temporal saliency algorithms respectively. From this, a hierarchical multi-scale model of a single person, his/her body pose and groups of people is formed. A person is represented by an elliptic blob where prominent oval-shaped parts are formed into a configuration. Interactions are identified by finding temporally correlated salient changes (we call events) in the probability distributions of our multi-scale configuration model. In this paper we will show how pose or configuration based models of the human body can provide a rich framework for modelling human-human interactive body behaviour even when body parts are occluded. In particular, the framework is suitable for extracting salient features from the human body where each part is represented by a few pixels in each image frame. The work is highly relevant to the development of automated systems for suspicious and antisocial behaviour detection and prevention.

Image fusion technology offers a range of potential benefits to the security and surveillance community, including increased situational awareness and enhanced detection capabilities. This paper reports on how image fusion technology is being adapted from the military arena (in which it is generally found today) to provide a new and powerful asset in the fight against crime and terrorism. The combination of detection and tracking processing with image fusion is discussed and assessed. Finally, a number of specific examples are presented which include land, air, and sea applications.

Accurate lip-reading techniques would be of enormous benefit for agencies involved in counter-terrorism and other law-enforcement areas. Unfortunately, there are very few skilled lip-readers, and it is apparently a difficult skill to transmit, so the area is under-resourced. In this paper we investigate the possibility of making the lip-reading task more amenable to a wider range of operators by enhancing lip movements in video sequences using active appearance models. These are generative, parametric models commonly used to track faces in images and video sequences. The parametric nature of the model allows a face in an image to be encoded in terms of a few tens of parameters, while the generative nature allows faces to be re-synthesised using the parameters. The aim of this study is to determine if exaggerating lip-motions in video sequences by amplifying the parameters of the model improves lip-reading ability. We also present results of lip-reading tests undertaken by experienced (but non-expert) adult subjects who claim to use lip-reading in their speech recognition process. The results, which are comparisons of word error-rates on unprocessed and processed video, are mixed. We find that there appears to be the potential to improve the word error rate but, for the method to improve the intelligibility there is need for more sophisticated tracking and visual modelling. Our technique can also act as an expression or visual gesture amplifier and so has applications to animation and the presentation of information via avatars or synthetic humans.

This paper reveals performance capabilities of state-of-the-art electro-optic and photonic devices, which are best suited for security and defense system applications. These devices can be used in battlefield, space surveillance, medical diagnosis, and detection of terrorist activities. Performance capabilities of fiber optic components for possible applications in WDM and DWDM systems are summarized. Photonic devices for covert military and security communication applications are identified with particular emphasis on performance and reliability. Performance parameters of Erbium-doped fiber amplifiers (EDFAs), Erbium-doped waveguide amplifiers (EDWAs), and optical hybrid amplifiers (OHAs) comprising of EDFAs and Raman amplifiers are discussed with emphasis on bandwidth, gain-flatness, data handling capability, channel capacity and cost-effectiveness. Performance parameters of very long wavelength infrared (VLWIR) detectors are summarized, which have potential applications in remote sensing and ballistic missile defense applications. Electro-optic and photonic devices best suited for security and defense applications are identified.

The market demand for bright laser pointers has led to the development of readily available devices that can pose a threat to road safety. Laser pointers can be involved in accidents caused by laser users who do not realise the dangers involved, but laser pointers can also enable deliberate criminal activity. There are technologies available that can counter the threat in different ways. A number of protective principles are outlined below. Some technologies built upon Liquid Crystal Devices are described in greater detail. Without any knowledge of what laser pointers a potential aggressor has access to, a frequency agile filter seems to be the most promising way to avoid the most severe consequences of dazzle from laser pointers. Protective systems incorporating suitable glasses or visors holding frequency agile filters of this kind however, are not commercially available today.

The huge amount of traffic transportable by the next generation optical network is vulnerable to attacks, as is discerned from an alarmingly increase of incidents. The types of attack are expected to range from the typical eavesdropping and service denial to more sophisticated source mimicking. As a consequence, modern encrypted methods refuge to highly sophisticated methods that emanate from quantum mechanics, known as quantum cryptography. However, the sophistication and elegance of quantum-cryptography makes the assumption that the transmission medium and the components involved on the link are perfect and that the properties of photons and the signal intensity do not change during propagation over many kilometers. Therefore, a practical implementation of Q-C may exhibit its own vulnerabilities due to non-linear interactions between photons and medium. Therefore, in addition to the sophistication of QKD and encryption algorithms, an additional function is needed that detects malicious intervention on the transmission link as well as a countermeasure strategy that outsmarts the attacker. In this paper, we consider a practical optical signal that consists of multiple photons, we consider a pragmatic medium with nonlinearities, scattering and absorption centers. We describe a case of service denial with Q-C, a method by which an attack is detected, and we develop a countermeasure strategy outsmarting the attacker. Our method assumes that the data channel is encrypted using sophisticated algorithms.

In this paper, we employ a simple theory based on driven damped oscillators to clarify the physical basis for message extraction in optical chaos communications using injection-locked semiconductor lasers. The receiver laser is optically driven by injection from the transmitter laser. We have numerically investigated the response characteristics of the receiver when it is driven by periodic (message) and chaotic (carrier) signals. It is thereby revealed that the response of the receiver laser in the two cases is quite different. For the periodic drive, the receiver exhibits a response depending on the signal frequency, while the chaotic drive provides a frequency-independent synchronous response to the receiver laser. CPF can be clearly understood in the difference between the periodic and chaotic drives. Message extraction using CPF is also examined, and the validity of our theoretical explanation for the physical mechanism underlying CPF is thus verified.

Optical visual cryptography based on binary computer generated hologram (BCGH) was proposed which used optics instead of human eyesight for decryption. As a result, it was possible to adapt cryptography to optical system. However, it also had some difficulties because it did not overcome the existing problem of visual cryptography completely. This paper suggested the method of optical cryptography implementation based on the phase modulation characteristics of liquid crystal display (LCD). The problems and appropriations are evaluated with some simulation. This system showed that the noise was reduced and resolution was improved compare with the conventional optical visual cryptography.

We demonstrate that, through coherent measurement of the transmitted terahertz frequency electric fields, broadband (0.3 - 8 THz) time-domain spectroscopy can be used to measure far-infrared vibrational modes of a range of drugs-of-abuse and high explosives that are of interest to the forensic and security services. Our results indicate that absorption features in these materials are highly sensitive to the structural and spatial arrangement of the molecules. Terahertz frequency spectra are also compared with high-resolution low-frequency Raman spectra to assist in understanding the low-frequency inter- and intra-molecular vibrational modes of the molecules.

We report a THz radiation source based on the technique of non-collinear phasematched parametric generation. The source, which is compact and operable at room temperature, generates nanosecond pulses of peak power and energy greater than 3 W and 17 nJ respectively. The radiation is continuously tunable over the range 1.2-3.05 THz and is of narrow spectral bandwidth (<100 GHz). The use of intersecting pump and parametric wave cavities results in minimum threshold pump pulse energies below 1 mJ (from a Nd:YAG laser excited at 20 W, 500 μsec by a quasi-CW diode-laser) and close to 50% down-conversion efficiency when operated at twice threshold.

There has been intense interest in the use of millimetre wave and terahertz technology for the detection of concealed weapons, explosives and other threats. Electromagnetic waves at these frequencies are safe, penetrate barriers and have short enough wavelengths to allow discrimination between objects. In addition, many solids including explosives have characteristic spectroscopic signatures at terahertz wavelengths which can be used to identify them. This paper reviews the progress which has been made in recent years and identifies the achievements, challenges and prospects for these technologies in checkpoint people screening, stand off detection of improvised explosive devices (IEDs) and suicide bombers as well as more specialized screening tasks.

Since the diffraction gratings were invented, their use in various security systems has been exploited. Their big advantage is the low production cost and, in the same time, the difficulty of replicating them. Most of the nowadays security systems are using those gratings to prove their originality. They can be seen on all the CDs, DVDs, most of the major credit cards and even on the wine bottles. In this article we present a new way of making such gratings without changing the production steps but generating an even more difficult to be replicated item. This new way consists not only in changing the grating period so that various false colours can be seen, but also their orientation so that for a complete check of the grating it should be seen under a certain solid angle. In the same time, one can also keep the possibility to change the grating period so this way various colours can be seen for each angle variation. By combining these two techniques (changing period and changing the angle ones) one can indeed create different images for each view angle and thus increasing the security of the object. In the same time, as can be seen, from the fabrication point of view no further complications appear. The production steps are identical, the only difference being the pattern. The resolution of the grating is not increased necessarily so neither from this point of view will complications appear.

Following Cascade Technologies first success at using Quantum Cascade Lasers (QCL) for trace gas detection in the continuous emission monitoring market, the core technology platform is now being developed towards homeland security applications. This paper will highlight the potential of QCL based trace gas sensor for detecting vapours of explosives. Furthermore we will present results that let foresee the use of such technologies at addressing security gaps for protection against terrorism in infrastructures where high throughput screening of individuals or items is required. Preliminary measurements have shown that rapid identification, or fingerprinting, of explosive is achievable in 10ms at extrapolated sensitivities in the sub-part per billion range. The experiments were carried out with support form the Home Office Scientific Development Branch (HOSDB) in the UK and were focused at selecting a variety of explosive compounds and showing their detection using a novel sniffer platform system based on the use of quantum cascade lasers. Preliminary studies on the technology have indicated that direct fingerprinting (detection - identification) of explosive compounds such as NG and tagging agents such as EGDN by sniffing surrounding ambient air is achievable. Furthermore these studies have also indicated that detection of such compounds on packaging used to ship the sealed compounds is possible, making this platform a strong contender for detection through cross contamination on material that have been in contact with each other. Additionally, it was also possible to detect breakdown products associated with sample material NG providing a further capability that could be exploited to enhance the detection and identification of explosive compounds.

Imaging techniques have been applied to a number of applications, such as translation and classification problems in medicine and defence. This paper examines the application of imaging techniques in digital forensics investigation using neural networks. A review of applications of digital image processing is presented, whiles a Pedagogical analysis of computer forensics is also highlighted. A data set describing selected images in different forms are used in the simulation and experimentation.

A proof-of-concept hand-held Raman spectrometer and a commercial portable system based on Attenuated Total Reflectance Fourier Transform Infrared spectroscopy (ATR-FTIR) were assessed for the rapid, "at scene" analysis of illicit drugs. The objectives of such an assessment were twofold: 1) to determine the suitability of the systems in practical forensic casework and 2) to determine the potential of the use of such systems in covert operations. Data obtained are promising and demonstrate the potential advantages and limitations of the use of these techniques in these fields of operation.

This presentation outlines the research we are developing in the area of Fourier Transform Infrared (FTIR) spectroscopic imaging with the focus on materials of forensic interest. FTIR spectroscopic imaging has recently emerged as a powerful tool for characterisation of heterogeneous materials. FTIR imaging relies on the ability of the military-developed infrared array detector to simultaneously measure spectra from thousands of different locations in a sample. Recently developed application of FTIR imaging using an ATR (Attenuated Total Reflection) mode has demonstrated the ability of this method to achieve spatial resolution beyond the diffraction limit of infrared light in air. Chemical visualisation with enhanced spatial resolution in micro-ATR mode broadens the range of materials studied with FTIR imaging with applications to pharmaceutical formulations or biological samples. Macro-ATR imaging has also been developed for chemical imaging analysis of large surface area samples and was applied to analyse the surface of human skin (e.g. finger), counterfeit tablets, textile materials (clothing), etc. This approach demonstrated the ability of this imaging method to detect trace materials attached to the surface of the skin. This may also prove as a valuable tool in detection of traces of explosives left or trapped on the surfaces of different materials. This FTIR imaging method is substantially superior to many of the other imaging methods due to inherent chemical specificity of infrared spectroscopy and fast acquisition times of this technique. Our preliminary data demonstrated that this methodology will provide the means to non-destructive detection method that could relate evidence to its source. This will be important in a wider crime prevention programme. In summary, intrinsic chemical specificity and enhanced visualising capability of FTIR spectroscopic imaging open a window of opportunities for counter-terrorism and crime-fighting, with applications ranging from analysis of trace evidence (e.g. in soil), tablets, drugs, fibres, tape explosives, biological samples to detection of gunshot residues and imaging of fingerprints.

A new laser-marked security label is presented. The label allows an integrated and distributed control scheme to be applied. The scheme incorporates a "closed circle" and allows internal control. A possible application for vehicles is discussed.

Surface enhanced resonance Raman scattering (SERRS) provides an increase in sensitivity over Raman scattering of about 10¹⁴ and rivals fluorescence in terms of its quantum efficiency. With the use of modern edge and notch filters and CCD cameras, the price and complexity of Raman spectroscopy equipment has decreased rapidly. This means that the potential advantages of SERRS are now much easier to release for use for practical purposes. The technique has specific advantages in terms of sensitivity and coding for use for tagging.

Conventional FTIR spectroscopy and microscopy has been widely used in forensic science. New opportunities exist to obtain rapid chemical images and to enhance the sensitivity of detection of trace materials using attenuated total reflection (ATR) Fourier transform infrared (FTIR) spectroscopy coupled with a focal-plane array (FPA) detector. In this work, the sensitivity of ATR-FTIR spectroscopic imaging using three different kinds of ATR crystals (Ge coupled with an infrared microscope, ZnSe and diamond) and resulting in three different optical arrangements for the detection of model drug particles is discussed. Model systems of ibuprofen and paracetamol particles having a size below 32 micrometers have been prepared by sieving. The sensitivity level in the three different approaches has been compared and it has been found that both micro and macro-ATR imaging methods have proven to be a promising techniques for the identification of concealed drug particles. To demonstrate the power and applicability of FTIR chemical imaging to forensic research, various examples are discussed. This includes investigation of the changes of chemical nature of latent fingerprint residue under controlled conditions of humidity and temperature studied by ATR-FTIR imaging. This study demonstrates the potential of spectroscopic imaging for visualizing the chemical changes of fingerprints.

Fingerprint recognition is widely used in the modern digital society and has become synonymous with reliable method of personal identification. However, artificial finger gummies with embedded fingerprints may easily spoof the current fingerprint system, which relies on the surface scanning of a finger. In this paper Optical Coherence Tomography (OCT) is applied to distinguish the artificial material, such as liquid silicone rubber and glue, as well as the real finger. Though the commercial fingerprint reader was easily spoofed by these artificial fingerprints, the OCT system easily detected them at all times.

Automatic face recognition is a useful tool in the fight against crime and terrorism. Technological advance in mobile communication systems and multi-application mobile devices enable the creation of hybrid platforms for active and passive surveillance. A dedicated mobile device that incorporates audio-visual sensors would not only complement existing networks of fixed surveillance devices (e.g. CCTV) but could also provide wide geographical coverage in almost any situation and anywhere. Such a device can hold a small portion of a law-enforcing agency biometric database that consist of audio and/or visual data of a number of suspects/wanted or missing persons who are expected to be in a local geographical area. This will assist law-enforcing officers on the ground in identifying persons whose biometric templates are downloaded onto their devices. Biometric data on the device can be regularly updated which will reduce the number of faces an officer has to remember. Such a dedicated device would act as an active/passive mobile surveillance unit that incorporate automatic identification. This paper is concerned with the feasibility of using wavelet-based face recognition schemes on such devices. The proposed schemes extend our recently developed face verification scheme for implementation on a currently available PDA. In particular we will investigate the use of a combination of wavelet frequency channels for multi-stream face recognition. We shall present experimental results on the performance of our proposed schemes for a number of publicly available face databases including a new AV database of videos recorded on a PDA.

The optical coherence tomography (OCT) is an emerging technology for high-resolution cross-sectional imaging of 3D structures. In the past years, OCT systems have been used mainly for medical, especially ophthalmological diagnostics. Concerning the nature of OCT system being capable to explore the internal features of an object, we apply the OCT technology to directly retrieve the 2D information pre-stored in a multiple-layer information carrier. The standard depth-resolution of an OCT system is at micrometer level. If a 20mm by 20mm sampling area with a 1024 x 1024 CCD array is used in the OCT system having 10 μm, an information carrier having a volume of 20mm x 20mm x 2mm could contain 200 Mega-pixel images. Because of its tiny size and large information volume, the information carrier, with its OCT retrieving system, will have potential applications in documents security and object identification. In addition, as the information carrier can be made by low-scattering transparent material, the signal/noise ratio will be improved dramatically. As a consequence, the specific hardware and complicated software can also be greatly simplified. Owing to non-scanning along X-Y axis, the full-field OCT could be the simplest and most economic imaging system for extracting information from such a multilayer information carrier. In this paper, deign and implementation of a full-field OCT system is described and the related algorithms are introduced. In our experiments, a four layers information carrier is used, which contains 4 layers of image pattern, two text images and two fingerprint images. The extracted tomography images of each layer are also provided.