The work described in this thesis, concerns developments to analytical microfluidic Lab-On-Chip platform originally developed by Prof Pamme's research group at the University of Hull. This work aims to move away from traditional laboratory analysis system towards a more effective system design which is fully automated and therefore potentially deployable in applications such as point of care medical diagnosis. The microfluidic chip platform comprises an external permanent magnet and chip with multiple parallel reagent streams through which magnetic micro-particles pass in sequence. These streams may include particles, analyte, fluorescent labels and wash solutions; together they facilitate an on-chip multi-step analytical procedure. Analyte concentration is measured via florescent intensity of the exiting micro-particles. This has previously been experimentally proven for more than one analytical procedure. The work described here has addressed a couple of issues which needed improvement, specifically optimizing the magnetic field and automating the measurement process. These topics are related by the fact that an optimal field will reduce anomalies such as aggregated particles which may degrade automated measurements. For this system, the optimal magnetic field is homogeneous gradient of sufficient strength to pull the particles across the width of the device during fluid transit of its length. To optimise the magnetic field, COMSOL (a Multiphysics simulation program) was used to evaluate a number of multiple magnet configurations and demonstrate an improved field profile. The simulation approach was validated against experimental data for the original single-magnet design. To analyse the results automatically, a software tool has been developed using C++ which takes image files generated during an experiment and outputs a calibration curve or specific measurement result. The process involves detection of the particles (using image segmentation) and object tracking. The intensity measurement follows the same procedure as the original manual approach, facilitating comparison, but also includes analysis of particle motion behaviour to allow automatic rejection of data from anomalous particles (e.g. stuck particles). For image segmentation a novel texture based technique called Temporal- Adaptive Median Binary Pattern (T-AMBP) combining with Three Frame Difference method to model the background for representing the foreground was proposed. This proposed approached is based on previously developed Adaptive Median Binary Pattern (AMBP) and Gaussian Mixture Model (GMM) approach for image segmentation. The proposed method successfully detects micro-particles even when they have very low fluorescent intensity, while most of the previous approaches failed and is more robust to noise and artefacts. For tracking the micro-particles, we proposed a novel algorithm called "Hybrid Meanshift", which combines Meanshift, Histogram of oriented gradients (HOG) matching and optical flow techniques. Kalman filter was also combined with it to make the tracking robust. The processing of an experimental data set for generating a calibration curve, getting effectively the same results in less than 5 minutes was demonstrated, without needing experimental experience, compared with at least 2 hours work by an experienced experimenter using the manual approach.

Computational imaging with single-pixel detectors utilises spatial correlation of light to obtain images. A series of structured illumination is generated using a spatial light modulator to encode the spatial information of an object. The encoded object images are recorded as total intensities with no spatial information by a single-pixel detector. These intensities are then sent to correlate with their corresponding illumination structures to derive an image. This correlation imaging method was first recognised as a quantum imaging technique called ghost imaging (GI) in 1995. Quantum GI uses the spatial correlation of entangled photon pairs to form images and was later demonstrated also by using classical correlated light beams. In 2008, an adaptive classical GI system called computational GI which employed a spatial light modulator and a single-pixel detector was proposed. Since its proposal, this computational imaging technique received intensive interest for this potential application. The aim of the work in this thesis was to improve this new imaging technique into a more applicable stage. Our contribution mainly includes three aspects. First an advanced reconstruction algorithm called normalised ghost imaging was developed to improve the correlation efficiency. By normalising the object intensity with a reference beam, the reconstruction single-to-noise ratio can be increased, especially for a more transmissive object. In the second work, a computational imaging scheme adapted from computational GI was designed by using a digital light projector for structured illumination. Compared to a conventional computational GI system, the adaptive system improved the reconstruction efficiency significantly. And for the first time, correlation imaging using structured illumination and single-pixel detection was able to image a 3 dimensional reflective object with reasonable details. By using several single-pixel detectors, the system was able to retrieve the 3 dimensional profile of the object. In the last work, effort was devoted to increase the reconstruction speed of the single-pixel imaging technique, and a fast computational imaging system was built up to generate real-time single-pixel videos.

With the development of pixelated liquid crystal displays, a new paradigm has emerged in the field of optics. Essentially, these displays enable interfacing a computer program with light, and therefore allow a wide range of light beams to be created. In this thesis, I shall be using liquid crystal displays to create phase diffraction patterns and, in this case, the displays are more commonly referred to as Spatial Light Modulators (SLMs). One area where SLMs have shown particular promise is that of optical microscopy. Here, they have been used in two different applications, namely holographic optical tweezers and SLM microscopy; this thesis concerns both. The aim of the thesis is to explore and develop new techniques combining SLMs with microscopy. The first part of this thesis goes into results of the experiments I have carried out in holographic optical tweezers. Hydrodynamic interactions play an important role in many physical and biological processes. I present experimental evidence for the partial synchronisation of the stochastic oscillations of two spheres in a bistable optical trap. This experiment showed that, even in the absence of an external driving force, a degree of synchronisation still exists due to the Brownian motion alone. I then describe a new procedure to protect the optical trap from contamination in sensitive samples. Microrheology using optical tweezers requires lengthy position measurements in order to obtain the linear viscoelastic properties of fluids and this measurement is often compromised by freely diffusing material entering the trap. I then apply rotational Doppler velocimetry to a particle spinning in an optical tweezers. This is the first time that structured illumination has been used to determine rotation rate in the micro regime. The second part describes the development of an SLM microscope and a series of experiments I carried out with it. The set up of the microscope is described and images are characterised in terms of the point spread function. I also demonstrate the multimodal capabilities by diffracting three different images, each with a unique spatial frequency filter, onto a single camera chip. Next, I report the development of some new frequency filters, namely holographic stereo microscopy and three variations, including stereo with defocus which mimics human binocular vision where we have two eyes (views) of the world, each having its own lens. I used 3D particle tracking to investigate sedimentation in a confined microscope sample. This experiment brought together SLM microscopy and optical tweezers to create a new technique for particle sizing, or study surface effects. This thesis describes several new applications of SLMs in microscopy, with the common theme being that the SLM is placed in the Fourier plane of the sample. Both holographic optical tweezers and SLM microscopy have been expanded by the techniques I have developed. In future, this work will serve as foundation for the combination for 3D particle tracking and visualisation with SLM microscopy, whilst microrheology will benefit from the new approaches.

Interferometry is used in a wide variety of fields for the instrumentation and analysis of subjects and the environment. When light beams interfere, an interference fringe pattern is generated. Captured widefield interference patterns can be used to determine changes in the optical path length of interfering beams across a 2D area. Two interferometer schemes regularly implemented in modern systems include the homodyne interferometer, where light with the same optical frequency in used to generate static intensity fringe patterns, and the heterodyne interferometer, where light with different optical frequencies are used to generate a fringe pattern that is modulated at a frequency equal to the optical frequency difference (beat frequency). A widefield heterodyne system is not straightforward to bring into practice, however, it does offer some benefits over a comparable homodyne interferometer, such as direct phase interpretation and the suppression of low frequency background light in interferograms. In this thesis, a widefield heterodyne interferometer system is presented. A custom prototype modulated light camera (MLC) chip was used to capture both homodyne and heterodyne fringe patterns. The 32x32 pixel camera is capable of continuously demodulating incident modulated light at frequencies between 100kHz and 17MHz. In the presented system, an error in the interferogram phase was determined to be Δφ = ±0.16radians (~9.1º). Comparisons between homodyne and heterodyne interferograms, captured using the MLC, are also presented. With modifications to the system, an ultrastable widefield heterodyne interferometer system was implemented. The intention of this system was to eliminate the contribution of piston phase to a captured interferogram without the need for common path optics. In contrast to the standard heterodyne setup, the reference signal used in the demodulation process was derived from one of the pixels on-board the MLC, rather than from an external source. This new local reference signal tracks the common changes in the temporal phase detected by all the MLC's pixels, eliminating piston phase and substantially reducing the contributions of unwanted vibrations and microphonics from interferograms. To demonstrate this ultrastable system, it is incorporated into a Mach-Zehnder interferometer, where a vibration is induced onto an object arm mirror (using a mounted speaker) at various frequencies. Stable interferograms are captured with the mirror moving at up to 85mm/s at 62Hz (an optical path length of 220μm, or 350 wavelengths for λ = 633nm), however, this limit was the result of the complex motion in the mirror mount rather than the stability limit of the system. The system is shown to be insensitive to pure piston phase variations equivalent to an object velocity of over 3m/s. As an application of the ultrastable system, a novel interferometer has been developed that captures the widefield fringe patterns generated by interfering two independent light sources, rather than by a single split source. The two separately stabilised HeNe lasers, constructed in the laboratory, produce light with a reasonably stable output frequency. Interfering two of these sources produce a heterodyne interference pattern with an unknown beat frequency. The beat frequency continuously varies because of the variation in the output frequency of each laser, but these stabilised lasers produce a beat frequency that drift by as little as 3MHz over 30 minutes. As the ultrastable system tracks changes in the temporal phase and instantaneous frequency of an incident fringe pattern, it can be used to track the variations in the modulation frequency generated by the fluctuations in the two separate lasers. The separation between the two lasers with regards to the images presented was about 35cm, but they can be separated by much larger amounts.

To obtain intrinsic reflectance of the scene by hyperspectral imaging systems has been a scientific and engineering challenge. Factors such as illumination variations, atmospheric effects and viewing geometries are common artefacts which modulate the way of light reflections from the object into the sensor and that they are needed to be corrected. Some of these factors induce highly scattered and diffuse irradiance which can artificially modify the intrinsic spectral reflectance of the surface, such as that in the shadows. This research is attempted to compensate the shadows in the hyperspectral imagery. In this study, three methods known as the Diffuse Irradiance Compensation (DIC), Linear Regression (LR) and spectro-polarimetry technique (SP) have been proposed to compensate the shadow effect. These methods have various degrees of shadow compensation capabilities, and their pros and cons are elucidated within the context of their classification performances over several data sets recorded within and outside of the laboratory. The spectro-polarimetry (SP) technique has been found to be the most versatile and powerful method for shadow compensation, and it has achieved over 90% of classification accuracy for the scenes with ~30% of shadow areas.

This thesis describes the development of a monolithic dual-wavelength laser based on an InP/AlGaInP quantum dot (QD) laser structure. Each wavelength is sourced from the same active region and can be operated simultaneously or independently, with light being emitted from a common aperture. The inhomogeneity of the QD material provides a wide distribution of energies, resulting in a broad and relatively flat-topped gain spectrum, which is ideal for sourcing multiple wavelengths. Measurements of optical absorption, gain and laser threshold current densities were used to characterise the optical properties of InP/AlGaInP QDs and ascertain a suitable structure from which to fabricate the dual-wavelength source. A growth temperature of 710 °C resulted in the lowest threshold current densities, and the incorporation of tensile strain into the upper confining layers was found to reduce the temperature dependence. Optical gain measurements were used to assess how state-filling and temperature govern the gain-peak wavelength. For a fixed gain at low injection the wavelength dependence follows that of the band gap (≈ 0.17 nm/K), but at higher levels of injection it becomes relatively temperature-insensitive. A minima in wavelength sensitivity corresponded to a net gain of ≈ 28 cm-1. Edge-emitting lasers with a wavelength temperature dependence as low as 0.03 nm/K were demonstrated for temperatures up to 107 °C (380 K). An Ar-Cl2 based inductively-coupled plasma (ICP) etch process, suitable for fabricating sub-micron features, was developed to create the necessary device architecture. Using the effects of state-filling and spectrally preferential feedback, coupled-cavity ridgewaveguide lasers with unequal length sections were used to generate two wavelengths, with separations up to 61.5 ± 0.2 nm. Time resolved spectra were used to demonstrate dual-mode operation, where both wavelengths were observed to emit simultaneously. This is a promising result as it suggests that this device could potentially be used as a compact terahertz source

Slow light is the study of the dramatic change in the velocity of light as it travels through certain media. This thesis focuses on slowing caused by transmitting light through a ruby crystal. When ruby experiences spatial or temporal modulation from a laser beam, the velocity of the light is greatly changed from its speed in a vacuum. The underlying mechanism for slow light in ruby is not fully understood and is, therefore, the subject of much debate. In this thesis, I examine many experimental parameters and their effects on slow light in ruby. First, I investigate the delay of images with both bright and dark regions through a spatial modulation of the ruby. I then turn to a temporal modulation of the ruby to answer the question of whether light can be delayed beyond the input pulse, the answer to which has the potential to differentiate between two proposed models for the mechanism that causes slow light in ruby. I return to the spatial domain to study the effect of spatial intensity distribution on the slow-light effect in ruby. I show that beams carrying orbital angular momentum rotate by an amount determined by the spatial feature of the beam profile. I present experimental evidence supporting a complicated mechanism of slow light in ruby, which informs the ongoing debate on the cause of slow light in ruby and provides direction for applications dependent upon the preservation of complex patterns in slow-light media.

No description available.

The rapid development of multimedia devices such as computers, network technologies, and cell phones have made it easier for users to create, broadcast, convey, share, store, and distribute multimedia data including images, videos and audio files on a daily basis. However, the availability of image processing software in the public domain has facilitated illegal copying and distribution of digital images with unnoticeable quality changes. Thus, security and identification of media content has become an important and demanding area for research. Perceptual hashing is one of the recent technologies used for multimedia content security. A perceptual image hash function is a hash function that is robust against content-preserving operations (CPOs), such as noise, JPEG lossy compression and rotation. This aim of this research is to study and investigate existing techniques and then contribute to the development of new perceptual image hashing techniques in the transform domain for image identification and copy detection applications. The design requirements for any perceptual image hashing system are robustness, discriminative capability (uniqueness), and unpredictability (security). The feature extraction stage plays a key role in ensuring the system output is robust and discriminative. This thesis mainly focuses on the robust feature extraction stage and the analysis of the proposed system's security. The following contributions have been made: A new perceptual hashing technique using pseudo-random sub-images in the discrete wavelet transform (DWT) domain for extracting features has been developed. The idea employs a recent dimension reduction technique, referred to as non-negative matrix factorization (NMF) in the literature, for enhancing the robustness and security of the hash. This approach is proposed to select the most stable coefficients under various content-preserving operations, compact. The robust image hashes are generated by applies DWT and NMF into image. The proposed sub-images-DWT technique has been shown to yield good performance under image processing operations, but it still suffers from geometric attacks. A new rotation-invariant FMT-based hashing technique incorporating the Fourier-Mellin transform and using overlapping blocks to improve the robustness against rotation attacks has also been proposed. The robust FMT-based image hashing is proposed to improve its performances under rotation, translation attacks and achieve better overall robustness. The invariance property to rotation, scaling and translation of FMT makes it more suitable for image hashing. Based on our experimental results, it has been shown that the proposed FMT-based image hashing technique is robust to a large class of image processing operations and geometric attacks. A new robust and secure DCT overlapping block-based hashing technique incorporating the discrete cosine transforms (DCT) to combat image processing attacks has been investigated. An improved DCT sign-based hashing technique robust against image processing attacks and well as small geometric manipulations developed. From the experimental results, it was observed that the low frequency coefficients for DCT sign based-image hashing were robust to a large of content-preserving operations (CPOs). The main idea was to exploit the energy compaction property of the DCT and its ability to carry information of edges and texture in DCT sign values. From the experimental results, it was observed that the low frequency coefficients for DCT sign-based image hashing were robust to a large class of content-preserving operations (CPOs). The main idea was to exploit the energy compaction property of the DCT and its ability to carry information of edges and texture in DCT sign values. Finally, the security of the proposed image hashing systems are discussed and analysed in the light of the corresponding design requirement. The DCT sign-based image hashing scheme has hash also been shown to be the most secure technique compared to other techniques proposed in this research as it offers the highest rate of bit independence in a hash.

Pas de résumé / No abstract