No description available.

No description available.

No description available.

The aim of this thesis is to track objects on a network of cameras both within [intra) and across (inter) cameras. The algorithms must be adaptable to change and are learnt in a scalable approach. Uncalibrated cameras are used that are patially separated, and therefore tracking must be able to cope with object oclusions, illuminations changes, and gaps between cameras.

A reliable method of remotely detecting concealed guns and explosives attached to the human body is of great interest to governments and security forces throughout the world. This thesis describes the development and trials of a new remote non-imaging concealed threat detection method using active millimetre wave radar using the microwave and mmwave frequencies bands 14 – 40 and 75 – 110 GHz (Ku, K, Ka and W). The method is capable of not only screening for concealed objects, like the current generation of concealed object detectors, but also of differentiating between mundane and threat objects. The areas focused upon during this investigation were: identifying the impact of different commonly worn fabrics as barriers to detection; consulting with end users about their requirements and operational needs; a comparison of different frequency bands for the detection of guns and explosives; exploring the effects of polarisation on object detection; a performance comparison of different detection schemes using Artificial Neural Networks; improving existing data acquisition systems and prototyping of a real-time capture system.

This thesis addresses the issues of applying advanced video analytics for surveillance applications. A video surveillance system can be defined as a technological tool that assists humans by providing an extended perception and capability of capturing interesting activities in the monitored scene. The prime components of video surveillance systems include moving object detection, object tracking, and anomaly detection. Moving object detection extracts the foreground silhouettes of moving objects. The object tracking component then applies the foreground information to create correspondences between tracks in the previous frame and objects in the current frame. The most challenging part of the system concerns the use of extracted scene information from the moving objects and object tracking for anomaly detection. The thesis proposes novel approaches for each of the main components above. They include: 1) an efficient foreground detection algorithm based on block-based detection and improved pixel-based Gaussian Mixture Model (GMM) refinement that can selectively update pixel information in each image region; 2) an adaptive object tracker that combines the merits of Kalman, mean-shift and particle filtering; 3) a feature clustering algorithm, which can automatically choose the optimal number of clusters in the training data for scene pattern classification; 4) a statistical scene modeller based on Bayesian theory and GMM, which combines object-based and local region-based information for enhanced anomaly detection. In addition, a layered feedback system architecture is proposed for using high- level detection results for improving low-level detection performance. Compared with common open-loop approaches, this increases the system reliability at the expense of using little extra computation. Moreover, considering the capability of real-time operation, robustness, and detection accuracy, which are key factors of video surveillance systems, appropriate trade-offs between complexity and detection performance are introduced in the relevant phases of the system, such as in moving object detection and in object tracking. The performance of the proposed system is evaluated with various video datasets. Both qualitative and quantitative measures are applied, for example visual comparison and precision-recall curves. The proposed moving object detection achieves an average of 52% and 38% improvement in terms of false positive detected pixels compared with a Gaussian Model (GM) and a GMM respectively. The object tracking component reduces the computation by 10% compared to a mean-shift filter while maintaining better tracking results. The proposed anomaly detection algorithm also outperforms previously proposed approaches. These results demonstrate the effectiveness of the proposed video surveillance system framework.

Person authentication is the process of confirming or determining a person's identity. Its purpose is to ensure that a system can only be accessed by authorised users. The Biometric method uses a person's physical or behavioural characteristics. The use of biometric characteristics is increasingly more popular as it makes unauthorised access more difficult.

This work is devoted to the study of the dynamic properties of magnetic amorphous wires, in particular, glass-coated microwires, which have small diameters (5-30 microns), outstanding soft magnetic behaviour with a high permeability and low coercivity, yet, possess a well-defined magnetic structure. First part of my PhD research has been devoted to the investigation of a bi-stable magnetisation reversal in glass-coated amorphous microwires. In contrast to traditional approaches, where characteristics of the magnetisation reversal are analysed as a consequence of the eddy current effect, l have applied stochastic methods for modelling the remagnetisation reversal in the microwires with axial anisotropy. While the eddy current approach, widely discussed in literature, was based on the single domain model, proposed stochastic approach takes into account a multi-domain state of studied samples. A modified stochastic Neel-Brown model of the magnetisation reversal has been proposed enabling the explanation of number of characteristic parameters of the microwires with axial magnetisation. Such important parameters of Barkhausen discontinuity as a mean switching field and a standard deviation of the switching field distribution have been investigated experimentally for understanding the influence of extrinsic factors such as a slew rate of the alternating magnetic field on applications operation. A deep understanding of the remagnetisation process in amorphous the microwires with axial anisotropy was successfully applied in development of a new type of the remote magnetic interrogation system. My reading system allows the large Barkhausen jump to be detected without actual contact between the magnetic microwire and the magnetic field detector. Experiments show that the detection will be possible at a distance of approximately 100-150 mm from the detecting sensor. A very low cost and easily repetitive amorphous microwires with axial anisotropy are . incontrovertibly best materials for Electronic Article Surveillance (EAS) applications. During the study of the microwires with axial anisotropy and development of the application based on them, I took part in the investigation of unusual coding methods of the amorphous microwires using a localised laser annealing treatment. This treatment produces a multi-pulse code within the wire and therefore adds to the information contained within the wire, improving reliability and security. I developed and used a magnetic interrogation system allowing an accurate and reliable test and analysis of the studied samples. The second part of my PhD research has included investigations of microwires with circumferential and helical anisotropies. The main interest in these materials is due to their applications for high-performance magnetic and stress sensors. Within this research project, the microwires with circumferential/helical anisotropy have been studied in a broad range of frequencies. A number of dynamic effects have been experimentally obtained and analysed. In particular, a detailed investigation of dynamic circular hysteresis (10kHz-300kHz) has been carried out allowing explanation of different behaviour of the materials with circumferential/helical anisotropy at different frequencies. The experimental curves are proposed to be analysed in terms of field dependence of characteristic permeabilities: domain wall displacements (reversible and irreversible) and magnetisation rotation. It was established that these permeabilities have different field behaviour. That explains different MI patterns at relatively low frequencies (less than a few MHz) and relatively high frequencies (more than 10 MHz). Further, some special features of the Magneto-Impedance effect in the microwires with a circumferential anisotropy such as off-diagonal impedance and microwave impedance have been considered. In this research, the former presents a considerable interest for development of linear magnetic sensors and the latter can find application in tuneable microwave materials and devices. As a result of this study several types of linear, bi-directional MI sensors were developed. I also developed new MI sensing approaches (such as off-diagonal response) and a new high performance detection technique allowing us to improve sensitivity, bandwidth, and linearity at low cost and simple construction. The last part of the PhD research has been devoted to an investigation of the stress-impedance in the ultra high-frequency (UHF) band (300MHz-3 GHz). Based on the experimental investigation, a new type of a stress-sensitive composite material is proposed. The microwave effective permittivity of such material depends on mechanical stresses. These composite materials opens up new possibilities for remote monitoring of stress with the use of microwave free-space techniques. This kind of composite material can be characterised as a 'sensing medium', which images the mechanical stress distribution inside construction or on its surface.

Designers of intrusion detection systems are often faced with the problem that their design fails to meet the specification because the actual implementation is not able to detect attacks as required. This work aims at addressing such shoncomings at an early stage of the design process. The proposed method provides guidance to intrusion detection systems designers by predicting whether or not a given design will be able to detect certain classes of attacks. Our method achieves this by introducing a classification of attacks and a description framework for intrusion detection systems. The attack classification and the description framework are defined at a common level of abstraction. and thereby form the basis for our analysis method which determines the attack classes that a given intrusion detection system design can detect. Intrusion detection system designers can use these results to determine where the design meets the specification and where it does not. These insights facilitate a more systematic and effective design process because they can be gained at an early stage of the design process without the need of actually implementing the design. Finally. we show how our approach to intrusion detection system design analysis can be validated and how the analysis results can be used for further applications such as guiding the design of intrusion detection architectures that combine diverse intrusion detection systems.

Multiview images are obtained by recording a scene from different viewpoints. The additional information can be used to improve the performance of various applications ranging from e-commerce to security surveillance. Many such applications process large arrays of images, and therefore it is important to consider how the information is stored and transmitted. In this thesis we address the issue of multiview image compression. Our approach is based on the concept that a point in a 3D space maps to a constant intensity line in specific multiview image arrays. We use this property to develop a sparse representation of multiview images. To obtain the representation we segment the data into layers, where each layer is defined by an object located at a constant depth in the scene. We extract the layers by initialising the layer contours and then by iteratively evolving them in the direction which minimises an appropriate cost function. To obtain the sparse representation we reduce the redundancy of each layer by using a multi-dimensional discrete wavelet transform (DWT). We apply the DWT in a separable approach; first across the camera viewpoint dimensions, followed by a 2D DWT applied to the spatial dimensions. The camera viewpoint DWT is modified to take into account the structure of each layer, and also the occluded regions. Based on the sparse representation, we propose two compression algorithms. The first is a centralised approach, which achieves a high compression, however requires the transmission of all the data. The second is an interactive method, which trades-off compression performance in order to facilitate random access to the multiview image dataset. In addition, we address the issue of rate allocation between encoding of the layer contours and the texture. We demonstrate that the proposed centralised and interactive methods outperform H.264/MVC and JPEG 2000, respectively.