This thesis considers sparse approximation of still images as the basis of a lossy compression system. The Matching Pursuit (MP) algorithm is presented as a method particularly suited for application in lossy scalable image coding. Its multichannel extension, capable of exploiting inter-channel correlations, is found to be an efficient way to represent colour data in RGB colour space. Known problems with MP, high computational complexity of encoding and dictionary design, are tackled by finding an appropriate partitioning of an image. The idea of performing MP in the spatio-frequency domain after transform such as Discrete Wavelet Transform (DWT) is explored. The main challenge, though, is to encode the image representation obtained after MP into a bit-stream. Novel approaches for encoding the atomic decomposition of a signal and colour amplitudes quantisation are proposed and evaluated. The image codec that has been built is capable of competing with scalable coders such as JPEG 2000 and SPIHT in terms of compression ratio.

The aim of the research is to improve laser system products manufactured at Selex ES which are used primarily for airborne targeting applications. This is achieved by developments to the design that prevent failures during manufacture or improve beam parameters such as divergence. A Q-switched diode-side-pumped Nd:YAG zig-zag geometry slab laser within a cross Porro prism resonator is investigated. This perturbation insensitive resonator design is used in laser systems operating over the full military environment of vibration and temperature. A number of aspects of the design are computer modelled with experimental verification, such as the effects of thermal lensing in the Nd:YAG slab, and the polarisation states in the resonator. These were used to analyse a number of issues encountered during manufacture, such as the lack of control over the polarisation state for output coupling, pre-lase causing damage to optical elements, and thermal lensing producing variations in beam quality. A number of design changes were made and, after experiments to verify improved performance, they were successfully integrated into a number of laser production programmes. The beam quality of laser systems was found to be affected by thermal lensing. A number of novel solutions were tested experimentally, which affected the thermal lens. Results of the alteration of the pump distribution in the Nd:YAG slab and the profile of conduction cooling are presented. 885 nm pumping instead of the traditional 808 nm pumping produced a reduction of the thermal lens by a factor of two from -0.1 D to -0.05 D, producing an improvement in the laser beam quality from M2 6.5 to 3.5. An enhancement in brightness of 2.2 was demonstrated using a laser resonator incorporating a deformable bimorph mirror. A new concept for a targeting laser source, which incorporated an eye-safe wavelength, was demonstrated using a common resonator intracavity OPO design. A conversion efficiency of 40% was achieved for 36 mJ output of 1573 nm eye-safe light from a 90 mJ laser at 1064 nm. The relative pointing directions of the two wavelength beams was measured to be within 250 μRad angular separation, which will be unaffected by ambient temperature variation. This level of performance is challenging to achieve in the current laser system design incorporating an extracavity OPO.

Laser target designators have been used for precision guidance purposes for many years. High energy laser devices are also being developed for tactical operations. Both are examples of Class 4 devices that can produce hazardous laser output. A triangular system-risk model links system performance, the operational demand and the potential hazards associated with the system operation, and which provides the basis for a system-level probabilistic risk assessment analysis. An extension of the risk assessment model includes contextual information as an additional risk element. A decomposition model, comprising a linear combination of irradiance basis vectors based on transverse Hermite-Gaussian modes, is developed using multiple linear regression modelling for the analysis of transient mode laser output. A linear predictive mixture distribution for a multiplicative scintillation gain factor is developed from simulated optical turbulence modelling. A spatial analysis using techniques from computational topology shows an approximately uniform distribution of high scintillation gain values over the sample surface pro le. A statistical analysis of level set critical points provides a more detailed characterisation of the scintillation gain variability. The combined decomposition and scintillation gain modelling provides a `subsystemlevel' performance analysis within the overall risk assessment context.

This thesis presents research results in two general areas of the solid-state laser technology commonly used in defence applications. The first area uses volume Bragg gratings (VBGs) as cavity mirrors, giving wavelength selection, linewidth reduction and improved frequency doubling efficiency. The second area is in extending the operating temperature range of diode-pumped Nd-doped pulsed lasers through the use of VBGs, external-cavity absorber-locked diode lasers and enhancements of pump chamber efficiency. For a laser-diode end-pumped Nd:YVO4 solid-state laser, a VBG laser cavity, with output powers of up to 6.2W is demonstrated at 1064 nm. With a conversion efficiency of 0.4 and M2 of 1.2, the laser linewidth is reduced by a factor of 16, to 72 pm, compared to a conventional dielectric mirror cavity. Intra-cavity losses due to the VBG are shown to be 2%. The 1064nm VBG cavity is then Q-switched using an AO modulator to generate peak powers of 2.3 kW. The second harmonic generation in a MgO:PPSLT crystal produced a 20% increase in second harmonic power, compared to a cavity with a conventional dielectric mirror. The improvement is attributable to the reduced linewidth. A similar configuration, but using a VBG with 98% reflectivity at 1342 nm, generates up to 3.8W of output power. The use of two VBG mirrors in a single cavity, further narrows the linewidth, allowing for output powers up to 2.3 W, in a single longitudinal mode. In-cavity heating of the VBG is found to give a reduction in reflectivity, a shift in the centre wavelength and increased linewidth. This is both modelled and experimentally demonstrated. For enhanced pumping of high-energy Q-switched Nd:YAG zigzag-slab lasers, two techniques of externally locking the wavelength of the QCW laser diodes are demonstrated. A chirped 808nm VBG locks the wavelength of a newly developed chirped quantum-well laser-diode array, without the need for fast-axis collimation. Locking over a temperature range of 50 C is demonstrated, with a predicted locking range of 70 C. As an alternative to VBG locking, a laser diode bar locked at 885nm is also demonstrated by using a thin Nd:YVO4 absorber. Filtered feedback locks the spectrum into peaks, which are in the Nd:YAG absorption bands over a 45 C range. In a parallel effort, a pump laser array configuration that increases the overall absorption path length from 1 to 3 cm in a side-pumped zigzag slab is demonstrated. By optimising the bounce lengths of the pump light and the inclusion of multi-wavelength laser bars, operation up to 80 C is demonstrated with an absorption efficiency of 80%. By adjusting the diode drive pulse duration, a Q-switched laser using this pump head operates from -40 to +60 C with a constant energy of 60 mJ. The operational temperature range of this configuration is shown to be over 120 C.

This thesis is dedicated to femtosecond combs as a tool for optical frequency metrology and as an integral part of an optical clock. After an overview of optical frequency measurement techniques, the design of two frequency combs based on mode-locked femtosecond lasers as they were at the beginning of my project is described. The first comb is based on an Er:fibre laser operating at a central wavelength of 1550 nm with a repetition rate of 100 MHz. The second is a Ti:sapphire-laser-based comb operating at a central wavelength of 810 nm with a repetition rate of 87 MHz. Improvements to the original design of the Ti:sapphire comb are detailed in the next chapter. A novel f-to-2f self-referencing scheme based on a pair of Wollaston prisms and employing a PPKTP crystal for SHG results in up to 20 dB enhancement of the signal to noise ratio in the carrier-envelope offset frequency beat signal f0 and in up to 15 dB lower phase noise in the f0 beat signal compared to a Michelson interferometer based system. Next, the factors influencing the stability and accuracy of the microwave reference signal and the performance of two synthesisers used for the stabilisation of the frequency combs were investigated. It is shown that stability of the maser reference signal is reduced by the distribution system by factor of 1.5. A fractional frequency change of 4.1(0.7) × 10−16 (K/h)−1 was measured for the better of the two synthesisers (an IFR 2023A) indicating that for accurate frequency measurements the synthesiser signal should be monitored to enable systematic frequency corrections to be made. Finally, an absolute frequency measurement of the electric quadrupole clock transition in a frequency standard based on a single 171Yb+ trapped ion is described. The result f = 688 358 979 309 310 ± 9 Hz agrees with an independent measurement made by the PTB group within the uncertainty of the measurements.

This thesis is concerned with the automatic detection and segmentation of visually salient image regions and subsequent targeted image compression in order to maintain observer performance levels while reducing image filesize. In moving towards this goal, pertinent issues have been addressed: the viability of "black-box" frequency transmission models, statistical measures of the effect of image processing, observer perception of processed images and how computer vision "feature points" correspond to visually salient image content. We show that image feature points are distributed towards visually-salient image regions: regions that are likely to attract observer attention. This remains true even when the "task" of the observer is changed: observers performing a task generally direct their attention towards image regions naturally rich in feature points. A new algorithm based on feature points, "Visual Interest", is proposed to predict image regions attended by observers. This method segments image content likely to attract visual attention under a variety of viewing conditions: passive viewing and search-directed viewing for different observer tasks. The algorithm improves the predictive power of observer eye fixations during object search task relative to "bottom-up" models. It responds only to image content, requiring no prior machine learning, in contrast to the scientific state-of-theart which relies explicitly on object categorisation. "Visual Interest" can also be run with object recognition to refine the segmentation for a particular object-category search task to reduce the "salient" area to tighter image areas. The resultant segmentation into salient and non-salient regions is used to generate regionof- interest compressed images suitable for multi-task observer analysis. Using pre-blur of JPEG we gain 15% filesize reduction beyond global JPEG application acting on image content alone and 25% when combined with object recognition. Using JPEG2000 ROI gives reductions of down to 25% of the original filesize while achieving gain in PSNR and SSIM statistics over the ROI, with the benefit of ROI priority transmission.

The work described in this thesis details the development of a multicore fibre device that can be used to optically trap multiple cells and particles. The optical trapping of multiple cells at close proximity allows for cell-to-cell interactions to be studied. Current methods available for creating arrays of traps are free space optical systems that use diffractive optics, laser scanning techniques or the interference of multiple beams to create the multiple traps. A fully integrated, fibre optic based, multiple particles, optical trapping device could be used in non-optical research facilities such as biological laboratories to aid with their research into cellular processes. In order to create the multiple traps, the distal end of the multicore fibre needs to be modified to induce a lensing effect. The multicore fibre device presented in this thesis was lensed in a fusion splicer; this refracts the outputs from the four cores to a common point in the far field where interference fringes are formed. The initial investigation demonstrated one-dimensional interferometric optical trapping through coupling light into two of the diagonal cores of the lensed multicore fibre. This produced linear interference fringes approximately 250 ± 25 μm from the end of the fibre with a fringe spacing of 2 ± 0.3 μm. The linear interference fringes were used to optically trap polystyrene microspheres with diameters of 1.3 μm, 2 μm and 3 μm in the high intensity regions of the fringes. Coupling into all four cores using a diffractive optical element produced an array of intensity peaks across the interference pattern with high visibility fringes greater than 80 %. Each intensity peak, spaced 2.75 μm apart could trap a single particle in two dimensions. The optical trapping of multiple microspheres and Escherichia coli bacterial cells was demonstrated proving that the lensed multicore fibre has the potential to be used to trap cells in biological experiments. The active manipulation of trapped 2 μm microspheres was also demonstrated through the rotation of the input polarisation to the multicore fibre. Finally, work towards creating a “turn-key” optical trapping device was demonstrated through the fabrication of a fully integrated multicore fibre device using an ultrafast laser-inscribed fan-out to couple light into each core. Single mode operation of the device was demonstrated at 1550 nm, using a weaker lensed MCF device. The two dimensional trapping of 4.5 μm polystyrene microspheres was shown in an array of peaks spaced 11.2 μm apart at a distance of 400 ± 25 μm from the end of the fibre.

This thesis contains the results of the work that was carried out between February 2009 and September 2012 in the area of high brightness light sources for defence applications. The work follows two main themes, namely nonlinearity in optical fibres and optical parametric oscillators (OPOs). Initially, the prospect of creating an ultrafast light source from solid core microstructured fibre via the phenomenon of modulation instability is discussed alongside supercontinuum generation and its value as a broad bandwidth source in countermeasures. What follows concerns OPOs in the infrared, both external (extracavity) and internal (intracavity) to a laser cavity, with the former is of benefit for high power applications, whereas the latter allows the OPO to operate more effectively at lower powers. The extracavity OPO section discusses a 2-stage conversion from 1064nm to 2128nm with the second stage design for conversion to ~5000nm, and the intracavity work is directed at both enabling a single frequency source for spectroscopy and examining the relationship between output coupling and the resonant OPO and laser fields.

Semiconductor disk lasers (SDLs) are efficient and commercially attractive devices as they can emit several Watts in continuous-wave (cw) operation, with good beam quality, low noise and great wavelength flexibility. SDL fundamental emission ranges from red to mid-infrared, while ultraviolet and other visible wavelength can be obtained via harmonic generation. This research shows that Raman conversion is an efficient way to extend the spectral coverage of well-established SDLs. The first experimental work consisted in the demonstration of a KGW Raman laser intracavity-pumped by a 1055 nm InGaAs SDL for laser emission at ~1.14 µm. This work represents the first Raman conversion of an SDL ever reported. Output power up to 0.8 W, broad wavelength tunability and cascaded Raman conversion (with low output coupling) were observed. The following experiment consisted in using synthetic single-crystal diamond as the Raman medium. In the last few years diamond has become a prominent Raman crystal as it provides high Raman gain, large Stokes shift and unrivalled thermal conductivity. A diamond Raman laser intracavity-pumped by an InGaAs SDL emitted up to 4.4 W at 1228 nm and was tuned from 1209-1256 nm. With an optical conversion efficiency exceeding 14%, this laser rivals the optical efficiencies of other cw Raman lasers and, perhaps more importantly, SDLs with direct emission at ~1.2 µm. Orange emission, with maximum output power of 1.5 W at 614 nm and wavelength tuning from 604.5-619.5 nm, was obtained via intracavity second harmonic generation in the Raman laser cavity. Raman conversion of a red-emitting GaInP SDL using a synthetic diamond crystal is also reported. As GaInP SDLs are less efficient than InGaAs SDLs, Raman threshold was more difficult to achieve, despite the Raman gain increasing for decreasing wavelengths. Nonetheless, Raman conversion in the deep red was observed, with output power of few tens of ~W, due to the low output coupling for the Raman laser, and tunable emission from 738-748 nm. This work is still at an early stage, so higher output power and enhanced laser efficiency may be achieved in future experiments.

Cellular contacts are crucial in determining function, yet these complex interactions are difficult to delineate due to their dynamic nature. This complexity is evident in the antigen-specific interactions between antigen-presenting cells (APCs, such as dendritic cells, B-cells or macrophages), and CD4+ T-cells. Recognition of cognate antigen by T-cells results in a rapid initiation of an intracellular signalling cascade, leading to translocation of specific transcription factors to the nucleus of the cell: factors that are important in determining the functional outcome of T-cell activation. These early interactions between T-cell and APC can determine the fate of a CD4+ T-cell. Factors such as quality, quantity, duration and strength of interaction between a T-cell and APC, can influence whether an effective immune response is initiated (and which type of response) or if a state of anergy is induced. Therefore understanding more about these factors that control this decision would help in the development of therapeutics that aim to initiate protective immunity or to improve selective suppression in autoimmunity and restore immune anergy. Within this thesis a novel approach is presented to dissect this interaction using optical tweezers. A novel optical tweezer setup is developed, providing greater control over cells and enhancing cell viability. This system is used to demonstrate quantification of the interactions between individual T cell and APC pairs, whereby the force of interaction increased from 3.3 (± 1.4) Piconewton (pN) in steady-state to 8.5 (± 5.7) pN upon antigen recognition. Importantly, the applicability of optical tweezers in addressing important biological questions was tested, investigating how T cell/APC interactions were altered during L-arginine deprivation, upon recognition of citrullinated antigen or in cells lacking an important signalling molecule. The approach presented here provides a novel tool for further understanding cell-cell interactions as well as demonstrating the potential for wide-ranging pharmaceutical screening applications.