Ce travail est dédié à la résolution de plusieurs problèmes de grand intérêt en traitement d’images : segmentation, choix de modèle et estimation de paramètres, pour le cas spécifique d’images texturées indirectement observées (convoluées et bruitées). Dans ce contexte, les contributions de cette thèse portent sur trois plans différents : modéle, méthode et algorithmique.Du point de vue modélisation de la texture, un nouveaumodèle non-gaussien est proposé. Ce modèle est défini dans le domaine de Fourier et consiste en un mélange de Gaussiennes avec une Densité Spectrale de Puissance paramétrique.Du point de vueméthodologique, la contribution est triple –troisméthodes Bayésiennes pour résoudre de manière :–optimale–non-supervisée–des problèmes inverses en imagerie dans le contexte d’images texturées ndirectement observées, problèmes pas abordés dans la littérature jusqu’à présent.Plus spécifiquement,1. la première méthode réalise la déconvolution myope non-supervisée et l’estimation des paramètres de la texture,2. la deuxième méthode est dédiée à la déconvolution non-supervisée, le choix de modèle et l’estimation des paramètres de la texture et, finalement,3. la troisième méthode déconvolue et segmente une image composée de plusieurs régions texturées, en estimant au même temps les hyperparamètres (niveau du signal et niveau du bruit) et les paramètres de chaque texture.La contribution sur le plan algorithmique est représentée par une nouvelle version rapide de l’algorithme Metropolis-Hastings. Cet algorithme est basé sur une loi de proposition directionnelle contenant le terme de la ”direction de Newton”. Ce terme permet une exploration rapide et efficace de l’espace des paramètres et, de ce fait, accélère la convergence. / This thesis is addressing a series of inverse problems of major importance in the fieldof image processing (image segmentation, model choice, parameter estimation, deconvolution)in the context of textured images. In all of the aforementioned problems theobservations are indirect, i.e., the textured images are affected by a blur and by noise. Thecontributions of this work belong to three main classes: modeling, methodological andalgorithmic. From the modeling standpoint, the contribution consists in the development of a newnon-Gaussian model for textures. The Fourier coefficients of the textured images are modeledby a Scale Mixture of Gaussians Random Field. The Power Spectral Density of thetexture has a parametric form, driven by a set of parameters that encode the texture characteristics.The methodological contribution is threefold and consists in solving three image processingproblems that have not been tackled so far in the context of indirect observationsof textured images. All the proposed methods are Bayesian and are based on the exploitingthe information encoded in the a posteriori law. The first method that is proposed is devotedto the myopic deconvolution of a textured image and the estimation of its parameters.The second method achieves joint model selection and model parameters estimation froman indirect observation of a textured image. Finally, the third method addresses the problemof joint deconvolution and segmentation of an image composed of several texturedregions, while estimating at the same time the parameters of each constituent texture.Last, but not least, the algorithmic contribution is represented by the development ofa new efficient version of the Metropolis Hastings algorithm, with a directional componentof the proposal function based on the”Newton direction” and the Fisher informationmatrix. This particular directional component allows for an efficient exploration of theparameter space and, consequently, increases the convergence speed of the algorithm.To summarize, this work presents a series of methods to solve three image processingproblems in the context of blurry and noisy textured images. Moreover, we present twoconnected contributions, one regarding the texture models andone meant to enhance theperformances of the samplers employed for all of the three methods.

This thesis begins by revisiting and building on themes of labour market segmentation, with particular reference given to Marx's seminal account of segmentation in Capital, Vol.1 (Chapter 25). Marx distinguishes between an active army - the stable full-time employed - and the relative surplus population - the precariously employed reserve army and the residual surplus - and suggests further fragmentation of these main groups into sub-strata. Marx's perspective of segmentation is grounded in fragments of a general theory of employment that, as a long-term tendency, identifies continual advances in constant capital that abolish work and proliferate the reserve army. This thesis builds on these themes by formulating a concept, the 'transference dynamic', which underpins a general theory of employment segmentation. A short history of segmentation under capitalism traces recent phases of development in both developed and lesser-developed nations. Stress is placed on the role of political configurations that regulate capitalism in ways that can either counter the general tendency, such is the case under the Fordist model of capitalism, or strengthen its logic. The theory of employment segmentation and the lessons drawn from the historical account are spliced together with an analysis of the contemporary phase of capitalism, labelled here as the neoliberal model of development. It is demonstrated that the coercive international regulatory dynamic of the neoliberal model reasserts and extends the competitive principle of the capitalist mode of production. Through this extension, nations are transformed into competition-states vying for scarce and globally mobile capital to operate on their shores - the primary source of national prosperity and employment - by implementing capital-friendly neoliberalized policy. This analysis of neoliberal global capitalism reveals an expanding surplus population within a context of deepening international segmentation. This employment crisis is expressed as a hierarchy of nations that is determined in part by their uneven development. Those at the bottom of the hierarchy, comprising a majority portion of the world's population, contain a massive reserve army and residual surplus population unincorporated into wage-based capitalism, without any obvious support of means of life and with little hope for the future. Finally, mainstream solutions are criticized for failing to address either long-run or contemporary drivers of the employment crisis. In response, this thesis pitches a project of multi-faceted radical reform that counter-regulates capitalism by adopting a combination of local, national, regional and global forms of democratic socialist governance.

Le sujet principal de cette thèse est la segmentation 3D de tumeurs du cerveau et de leurs différentes composantes (oedème et nécrose), ainsi que de structures internes du cerveau en IRM. Pour la segmentation de tumeurs nous proposons un cadre général qui est une combinaison des paradigmes fondés sur les régions et les contours. Dans ce cadre, nous segmentons d'abord le cerveau en utilisant une méthode adaptée aux cas pathologiques et extrayons des informations globales sur la tumeur par analyse de symétrie. La deuxième étape segmente la tumeur et ses composantes. Pour cela, nous proposons une méthode nouvelle et originale qui combine l'information de régions et de contours en deux phases. Pour la première, l'initialisation, nous présentons deux nouvelles méthodes. La première est une nouvelle méthode de classification floue qui exploite à la fois l'information des voxels et leurs voisinages (inspirés des champs Markov (MRF)), l'appartenance et la typicalité. La seconde se fonde sur l'analyse de la symétrie. La segmentation initiale de la tumeur est raffinée dans la deuxième phase par un modèle déformable contraint par des relations spatiales. Les relations spatiales sont obtenues en utilisant la segmentation initiale et les tissus environnant la tumeur. La méthode proposée peut être employée pour une grande classe de tumeurs dans n'importe quelle modalité en IRM. Pour segmenter une tumeur et ses composantes automatiquement, le cadre proposé a besoin seulement d'une image CE-T1w (con- trast enhanced T1-weighted) et d'une image FLAIR. Dans le cas d'une image CE-T1w seulement, l'interaction de l'utilisateur peut être nécessaire. Nous avons évalué cette méthode sur une base de données de 20 images CE-T1w et 10 images FLAIR avec différents types de tumeurs. Un autre but de cette thèse est la segmentation de structures internes du cerveau en présence d'une tumeur. Pour cela, une connaissance a priori sur l'anatomie et l'organisation spatiale des structures est fournie par une ontologie. Pour segmenter chaque structure, nous exploitons ses relations spatiales par rapport à d'autres structures, selon la connaissance a priori. Nous choisissons alors les relations spatiales qui sont valables en fonction de la tumeur segmentée. Ces relations spatiales sont alors modélisées dans un cadre flou proposé par notre groupe. Comme pour la tumeur, la procédure de segmentation de chaque structure comporte deux étapes. Dans la première étape nous recherchons la segmentation initiale de la structure dans le cerveau globalement segmenté. Le processus de recherche est fait dans la région d'intérêt fournie par la fusion des relations spatiales. Pour segmenter globalement les structures du cerveau nous employons deux méthodes. La première est la classification floue propos ée et la seconde repose sur les ensembles de niveaux multi-phases. Pour raffiner la segmentation initiale, nous employons un modèle déformable qui est contraint par les relations spatiales de la structure. Cette méthode a été également évaluée sur 10 images CE-T1w pour segmenter les ventricules, les noyaux caudés et les thalami.

Thesis (M.A. (Tourism))--North-West University, Potchefstroom Campus, 2009.

The haptic exploration of 2-D images is a challenging problem in computer haptics. Research on the topic has primarily been focused on the exploration of maps and curves. This thesis describes the design and implementation of a system for the haptic exploration of photographs. The system builds on various research directions related to assistive technology, computer haptics, and image segmentation. An object-level segmentation hierarchy is generated from the source photograph to be rendered haptically as a contour image at multiple levels-of-detail. A tool for the authoring of object-level hierarchies was developed, as well as an innovative type of user interaction by region selection for accurate and efficient image segmentation. According to an objective benchmark measuring how the new method compares with other interactive image segmentation algorithms shows that our region selection interaction is a viable alternative to marker-based interaction. The hierarchy authoring tool combined with precise algorithms for image segmentation can build contour images of the quality necessary for the images to be understood by touch with our system. The system was evaluated with a user study of 24 sighted participants divided in different groups. The first part of the study had participants explore images using haptics and answer questions about them. The second part of the study asked the participants to identify images visually after haptic exploration. Results show that using a segmentation hierarchy supporting multiple levels-of-detail of the same image is beneficial to haptic exploration. As the system gains maturity, it is our goal to make it available to blind users.

During vertebrate embryo development, the body axis is subdivided into repeated structures, called somites. Somites bud off from an un-segmented tissue called the pre-somitic mesoderm (PSM) in a sequential and periodic manner, tightly controlled by an in built molecular clock, called the "segmentation clock". According to current understanding, the clock is comprised of: (i) an autonomous cellular oscillator consisting of an intracellular negative feedback loop of Her genes within the PSM cells, (ii) Delta-ligand and Notch-receptor coupling that facilitates synchronization of oscillators among the PSM cells, (iii) Tissue level waves of gene expression that emerge in the posterior PSM and move coherently towards anterior, leading to global arrest of oscillations in the form of somites. However, the movement of cellular oscillators within the PSM before the formation of somitic furrows, a prominent feature of the tissue as observed through this work has not been experimentally considered as a constituent of the segmentation clock so far. Our work aims to incorporate movement of cellular oscillators in the framework of the segmentation clock. It is well known from theoretical studies that the characteristics of relative motion of oscillators affect their synchronization properties and the patterns of oscillations they form. Particularly, theoretical studies by Uriu et al., PNAS (2010) suggest that cell movements promotes synchronization of genetic oscillations. Here, we established experimental techniques and image analysis tools to attain quantitative insight on (i) diffusion co-efficient of cellular oscillators, (ii) dynamics of a population of oscillators, within the PSM aiming towards concomitant understanding of the relationship between movement and synchronization of cellular oscillators. In order to quantitatively relate cellular oscillators and their motion within the PSM, I established imaging techniques that enabled visualization of fluorescenctly labeled nuclei as readouts of cell positions in live embryo, and developed dedicated segmentation algorithm and implemented tracking protocol to obtain nuclei positions over time in 3D space. Furthermore, I provide benchmarking techniques in the form of artificial data that validate segmentation algorithm efficacy and, for the first time proposed the use of transgenic embryo chimeras to validate segmentation algorithm performance within the context of in vivo live imaging of embryonic tissues. Preliminary analysis of our data suggests that there is relatively high cell mixing in the posterior PSM, within the same spatial zone where synchronous oscillations emerge at maximum speed. Also, there are indications of gradient of cell mixing along the anterior-posterior axis of the embryo. By sampling single cell tracks with the help of nuclei markers, we have also been able to follow in vivo protein oscillations at single cell resolution that would allow quantitative characterization of coherence among a population of cellular oscillators over time. Our image analysis work flow allows testing of mutant embryos and perturbation of synchrony dynamics to understand the cause-effect relationship between movement and synchronization properties at cellular resolution. Essentially, through this work, we hope to bridge the time scales of events and cellular level dynamics that leads to highly coordinated tissue level patterns and thereby further our understanding of the segmentation clock mechanism.

The haptic exploration of 2-D images is a challenging problem in computer haptics. Research on the topic has primarily been focused on the exploration of maps and curves. This thesis describes the design and implementation of a system for the haptic exploration of photographs. The system builds on various research directions related to assistive technology, computer haptics, and image segmentation. An object-level segmentation hierarchy is generated from the source photograph to be rendered haptically as a contour image at multiple levels-of-detail. A tool for the authoring of object-level hierarchies was developed, as well as an innovative type of user interaction by region selection for accurate and efficient image segmentation. According to an objective benchmark measuring how the new method compares with other interactive image segmentation algorithms shows that our region selection interaction is a viable alternative to marker-based interaction. The hierarchy authoring tool combined with precise algorithms for image segmentation can build contour images of the quality necessary for the images to be understood by touch with our system. The system was evaluated with a user study of 24 sighted participants divided in different groups. The first part of the study had participants explore images using haptics and answer questions about them. The second part of the study asked the participants to identify images visually after haptic exploration. Results show that using a segmentation hierarchy supporting multiple levels-of-detail of the same image is beneficial to haptic exploration. As the system gains maturity, it is our goal to make it available to blind users.

Lung carcinoma is the most prevalent type of cancer in the world, considered to be a relentlessly progressive disease, with dismal mortality rates to patients. Recent advances in targeted therapy hold the premise for the delivery of better, more effective treatments to lung cancer patients, that could significantly enhance their survival rates. Optimizing care delivery through targeted therapies requires the ability to effectively identify and diagnose lung cancer along with identifying the lung cancer cell type specific to each patient, \textit{small cell carcinoma}, \textit{adenocarcinoma}, or \textit{squamous cell carcinoma}. Label free optical imaging techniques such as the \textit{Coherent anti-stokes Raman Scattering microscopy} have the potential to provide physicians with minimally invasive access to lung tumor sites, and thus allow for better cancer diagnosis and sub-typing. To maximize the benefits of such novel imaging techniques in enhancing cancer treatment, the development of new data analysis methods that can rapidly and accurately analyze the new types of data provided through them is essential. Recent studies have gone a long way to achieving those goals but still face some significant bottlenecks hindering the ability to fully exploit the diagnostic potential of CARS images, namely, the streamlining of the diagnosis process was hindered by the lack of ability to automatically detect cancer cells, and the inability to reliably classify them into their respective cell types. More specifically, data analysis methods have thus far been incapable of correctly identifying and differentiating the different non-small cel lung carcinoma cell types, a stringent requirement for optimal therapy delivery. In this study we have addressed the two bottlenecks named above, through designing an image processing framework that is capable of, automatically and accuratly, detecting cancer cells in two and three dimensional CARS images. Moreover, we built upon this capability with a new approach at analyzing the segmented data, that provided significant information about the cancerous tissue and ultimately allowed for the automatic differential classification of non-small cell lung carcinoma cell types, with superb accuracies.

Speech Endpoint Detection, also known as Speech Segmentation, is an unsolved problem in speech processing that affects numerous applications including robust speech recognition. This task is not as trivial as it appears, and most of the existing algorithms degrade at low signal-to-noise ratios (SNRs). Most of the previous research approaches have focused on the development of robust algorithms with special attention being paid to the derivation and study of noise robust features and decision rules. This research tackles the endpoint detection problem in a different way, and proposes a novel speech endpoint detection algorithm which has been derived from Chan-Vese algorithm for image segmentation. The proposed algorithm has the ability to fuse multi features extracted from the speech signal to enhance the detection accuracy. The algorithm performance has been evaluated and compared to two widely used speech detection algorithms under various noise environments with SNR levels ranging from 0 dB to 30 dB. Furthermore, the proposed algorithm has also been applied to different types of American English phonemes. The experiments show that, even under conditions of severe noise contamination, the proposed algorithm is more efficient as compared to the reference algorithms.

Abstract The purpose of this report was to study if market segmentation could influence the corporate identity and indirectly the image of a company. The relevance of this purpose was due to a lack of development of the market segmentation theory in a long period of time. There are many authors in this area of research, but not much of a consensus between them. This report therefore delves deeper into one of these market segmentation theories in order to verify the multi-segmentation theory.   During the process of investigation the management of four music festivals were interviewed to see how they use their market segmentation. Several interviews were done with one or more individuals of the organizations. The findings are two-folded. The use of multi-segmentation was proven through the interviews with the festivals. All four festivals used several different market segmentation variables, all focusing greater understanding of the values and personalities of their target group.   The most interesting finding is that from the greater understanding of their customers, the festivals have been able to identify an identity with their target group. Their respective corporate identities have then been build around their target group identities in order to reflect the customers values and believes. This concept of a target group identity may have great impacts on the customers’ perception of the corporate identity and image.