This dissertation presents algorithms for restoring some of the major corruptions observed in archived film or video material. The two principal problems of impulsive distortion (Dirt and Sparkle or Blotches) and noise degradation are considered. There is also an algorithm for suppressing the inter-line jitter common in images decoded from noisy video signals. In the case of noise reduction and Blotch removal the thesis considers image sequences to be three dimensional signals involving evolution of features in time and space. This is necessary if any process presented is to show an improvement over standard two-dimensional techniques. It is important to recognize that consideration of image sequences must involve an appreciation of the problems incurred by the motion of objects in the scene. The most obvious implication is that due to motion, useful three dimensional processing does not necessarily proceed in a direction 'orthogonal' to the image frames. Therefore, attention is given to discussing motion estimation as it is used for image sequence processing. Some discussion is given to image sequence models and the 3D Autoregressive model is investigated. A multiresolution BM scheme is used for motion estimation throughout the major part of the thesis. Impulsive noise removal in image processing has been traditionally achieved by the use of median filter structures. A new three dimensional multilevel median structure is presented in this work with the additional use of a detector which limits the distortion caused by the filters . This technique is found to be extremely effective in practice and is an alternative to the traditional global median operation. The new median filter is shown to be superior to those previously presented with respect to the ability to reject the kind of distortion found in practice. A model based technique using the 3D AR model is also developed for detecting and removing Blotches. This technique achieves better fidelity at the expense of heavier computational load. Motion compensated 3D IIR and FIR Wiener filters are investigated with respect to their ability to reject noise in an image sequence. They are compared to several algorithms previously presented which are purely temporal in nature. The filters presented are found to be effective and compare favourably to the other algorithms. The 3D filtering process is superior to the purely temporal process as expected. The algorithm that is presented for suppressing inter-line jitter uses a 2D AR model to estimate and correct the relative displacements between the lines. The output image is much more satisfactory to the observer although in a severe case some drift of image features is to be expected. A suggestion for removing this drift is presented in the conclusions. There are several remaining problems in moving video. In particular, line scratches and picture shake/roll. Line scratches cannot be detected successfully by the detectors presented and so cannot be removed efficiently. Suppressing shake and roll involves compensating the entire frame for motion and there is a need to separate global from local motion. These difficulties provide ample opportunity for further research.

Tato diplomová práce nastiňuje teoretické základy potřebné pro návrh algoritmu rekurzivní metody nejmenších čtverců a následně jeho aplikaci na experimentální data naměřená při testovacím manévru uskutečněném v roce 2001. Analyzována byla příčná dynamika jednostopého rovinného modelu vozidla. Práce také obsahuje srovnání výsledků získaných jednak rekurzivním algoritmem a dále i algoritmem Kalmanova filtru.

<p>Doktorska disertacija je posvećena re&scaron;avanju nelinearnih diferen&shy; cijalnih jednačina, kao i linearnih diferencijalnih jednačina sa singularite-tim a u okviru prostora Kolomboovih uop&scaron;tenih funkcija. U osnovi, dis&shy;ertacija se može podeliti na dva dela. Prvi deo disertacije je posvećen re&scaron;avanju determinističkih parcijalnih diferencijalnih jednačina primenom teorije polugrupa operatora definisanih na prostorima Kolomboa. Drugi deo disertacije posvećen je re&scaron;avanju stohastičkih običnih i parcijalnih dife&shy;rencijalnih jednačina. Ove jednačine sadrže Kolomboove uop&scaron;tene slučajne procese kao nelinearni deo, ili kao početne uslove.</p> / <p>Doctoral thesis is devoted to nonlinear differential equations, as well as linear differential equations with singularities in the framework of Colombeau generalized function spaces. Basically, the thesis can be devided into two parts. The first part is devoted to solving deterministic partial differential equations applaying semigroup theory where those semigroups are defined on Colombeau spaces. The second part of the thesis is devoted to stochastic ordinary and partial differential equations. Those equations contain Colombeau generalized stochastic processes as nonlinear part, or as initial data.</p>

Nous proposons une étude de l’utilisation avancée de l’hypothèse de blancheur du bruit pour améliorer les performances de débruitage. Nous mettons en avant l’intérêt d’évaluer la blancheur du résidu par des mesures de corrélation dans différents cadres applicatifs. Dans un premier temps, nous nous plaçons dans un cadre variationnel et nous montrons qu’un terme de contrainte sur la blancheur du résidu peut remplacer l’attache aux données L2 en améliorant significativement les performances de débruitage. Nous le complétons ensuite par des termes de contrôle de la distribution du résidu au moyen des moments bruts. Dans une seconde partie, nous proposons une alternative au rapport de vraisemblance menant, à la norme L2 dans le cas Gaussien blanc, pour mesurer la dissimilarité entre patchs. La métrique introduite, fondée sur l’autocorrélation de la différence des patchs, se révèle plus performante pour le débruitage et la reconnaissance de patchs similaires. Finalement, les problématiques d’évaluation de qualité sans oracle et de choix local de modèle sont abordées. Encore une fois, la mesure de la blancheur du résidu apporte une information pertinente pour estimer localement la fidélité du débruitage. / We propose an advanced use of the whiteness hypothesis on the noise to imrove denoising performances. We show the interest of evaluating the residual whiteness by correlation measures in multiple applications. First, in a variational denoising framework, we show that a cost function locally constraining the residual whiteness can replace the L2 norm commonly used in the white Gaussian case, while significantly improving the denoising performances. This term is then completed by cost function constraining the residual raw moments which are a mean to control the residual distribution. In the second part of our work, we propose an alternative to the likelihood ratio, leading to the L2 norm in the white Gaussian case, to evaluate the dissimilarity between noisy patches. The introduced metric, based on the autocorrelation of the patches difference, achieves better performances both for denoising and similar patches recognition. Finally, we tackle the no reference quality evaluation and the local model choice problems. Once again, the residual whiteness bring a meaningful information to locally estimate the truthfulness of the denoising.

Ce travail, à la frontière de l’analyse et des probabilités, s’intéresse à l’étude de systèmes différentiels a priori mal posés. Nous cherchons, grâce à des techniques issues de la théorie des chemins rugueux et de l’étude trajectorielle des processus stochastiques, à donner un sens à de tels systèmes puis à les résoudre, tout en montrant que les notions proposées ici étendent bien les notions classiques de solutions. Cette thèse se décompose en trois chapitres. Le premier traite des systèmes différentiels ordinaires perturbés additivement par des processus irréguliers éventuellement stochastiques ainsi que des effets de régularisation de tels processus. Le deuxième chapitre concerne l’équation de transport linéaire perturbée multiplicativement par des chemins rugueux ; enfin, le dernier chapitre s’intéresse à une équation de la chaleur non linéaire perturbée par un bruit blanc espace-temps, l’équation de quantisation stochastique phi4 en dimension 3. / In this work we investigate a priori ill-posed differential systems from an analytic and probabilistic point of view. Thanks to technics inspired by the rough path theory and pathwise study of stochastic processes, we want to define those ill-posed systems and then study them. The first chapter of this thesis is related to ordinary differential equations perturbed by some irregular (stochastic) processes and the effects induced by the regularization of such processes. The second chapter deals with the linear transport equation multiplicatively perturbed by a rough path. Finally, in the last chapter we investigate the stochastic quantization equation Phi4 in three dimensions.

Artificial Neural Networks (ANNs) can be viewed as a mathematical model to simulate natural and biological systems on the basis of mimicking the information processing methods in the human brain. The capability of current ANNs only focuses on approximating arbitrary deterministic input-output mappings. However, these ANNs do not adequately represent the variability which is observed in the systems' natural settings as well as capture the complexity of the whole system behaviour. This thesis addresses the development of a new class of neural networks called Stochastic Neural Networks (SNNs) in order to simulate internal stochastic properties of systems. Developing a suitable mathematical model for SNNs is based on canonical representation of stochastic processes or systems by means of Karhunen-Loéve Theorem. Some successful real examples, such as analysis of full displacement field of wood in compression, confirm the validity of the proposed neural networks. Furthermore, analysis of internal workings of SNNs provides an in-depth view on the operation of SNNs that help to gain a better understanding of the simulation of stochastic processes by SNNs.

La vitesse des liens séries haut débit (USB, SATA, PCI-express, etc.) a atteint les multi-gigabits par seconde, et continue à augmenter. Deux des principaux paramètres électriques utilisés pour caractériser les performances des SerDes sont la gigue transmis à un niveau de taux d’erreur donné et la capacité du récepteur à suivre la gigue à un taux d’erreur donné.Modéliser le bruit de phase des différents components du SerDes, et extraire la gigue temporelle pour la décomposer, aideraient les ingénieurs en conception de circuits à atteindre les meilleurs résultats pour les futures versions des SerDes. Générer des patterns de gigue synthétiques de bruits blancs ou colorés permettrait de mieux analyser les effets de la gigue dans le système pendant la phase de vérification.La boucle d’asservissement de phase est un des contributeurs de la gigue d’horloge aléatoire et déterministe à l’intérieur du système. Cette thèse présente une méthode pour modéliser la boucle d’asservissement de phase avec injection du bruit de phase et estimation de la gigue temporelle. Un modèle dans le domaine temporel en incluant les effets de non-linéarité de la boucle a été créé pour estimer cette gigue. Une nouvelle méthode pour générer des patterns synthétiques de gigue avec une distribution Gaussienne à partir de profils de bruit de phase coloré a été proposée.Les standards spécifient des budgets séparés de gigue aléatoire et déterministe. Pour décomposer la gigue de la sortie de la boucle d’asservissement de phase (ou la gigue généré par la méthode présentée), une nouvelle technique pour analyser et décomposer la gigue a été proposée. Les résultats de modélisation corrèlent bien avec les mesures et cette technique aidera les ingénieurs de conception à identifier et quantifier proprement les sources de la gigue ainsi que leurs impacts dans les systèmes SerDes.Nous avons développé une méthode, pour spécifier la boucle d’asservissement de phase en termes de bruit de phase. Cette méthode est applicable à tout standard (USB, SATA, PCIe, …) et définit les profils de bruits de4phases pour les différentes parties de la boucle d’asservissement de phase, pour s’assurer que les requis des standards sont satisfaits en termes de gigue. Ces modèles nous ont également permis de générer les spécifications de la PLL, pour des standards différents. / Bit rates of high speed serial links (USB, SATA, PCI-express, etc.) have reached the multi-gigabits per second, and continue to increase. Two of the major electrical parameters used to characterize SerDes Integrated Circuit performance are the transmitted jitter at a given bit error rate (BER) and the receiver capacity to track jitter at a given BER.Modeling the phase noise of the different SerDes components, extracting the time jitter and decomposing it, would help designers to achieve desired Figure of Merit (FoM) for future SerDes versions. Generating white and colored noise synthetic jitter patterns would allow to better analyze the effect of jitter in a system for design verification.The phase locked loop (PLL) is one of the contributors of clock random and periodic jitter inside the system. This thesis presents a method for modeling the PLL with phase noise injection and estimating the time domain jitter. A time domain model including PLL loop nonlinearities is created in order to estimate jitter. A novel method for generating Gaussian distribution synthetic jitter patterns from colored noise profiles is also proposed.The Standard Organizations specify random and deterministic jitter budgets. In order to decompose the PLL output jitter (or the generated jitter from the proposed method), a new technique for jitter analysis and decomposition is proposed. Modeling simulation results correlate well with measurements and this technique will help designers to properly identify and quantify the sources of deterministic jitter and their impact on the SerDes system.We have developed a method, for specifying PLLs in terms of Phase Noise. This method works for any standard (USB, SATA, PCIe, …), and defines Phase noise profiles of the different parts of the PLL, in order to be sure that the standard requirements are satisfied in terms of Jitter.

Les possibles effets bénéfiques de perturbations déterministes ou stochastiques sur la réponse de différents systèmes non linéaires sont étudiés. À cet effet, des études numériques et expérimentales sont conjointement proposées sur deux structures distinctes : un oscillateur électronique de type FitzHugh-Nagumo et une ligne électrique constituée de 45 de ces oscillateurs couplés résistivement. La caractérisation de l’oscillateur élémentaire est d’abord réalisée en régime déterministe. En présence d’une excitation bichromatique, il est notamment montré que lorsque la composante de fréquence la plus faible est subliminale, sa détection en sortie du système peut être maximisée pour une amplitude particulière de la seconde composante, qui agit alors comme une perturbation haute fréquence. Par la suite, il est établi que ce phénomène de résonance vibrationnelle peut être amélioré pour quelques fréquences spécifiques de cette perturbation déterministe. Par ailleurs, en introduisant une composante stochastique dans l’excitation, l’attention est ensuite portée sur le phénomène de résonance stochastique fantôme. Celui-ci se distingue par le fait que la fréquence d’intérêt en sortie du système ne fait désormais plus partie du signal excitateur. La dernière partie est consacrée à l’étude de la structure couplée. Il est montré que la propagation d’une information à travers les cellules de la ligne peut être améliorée via les phénomènes de propagation vibrationnelle et de propagation assistée par le bruit. Ceux-ci se produisent sous certaines conditions, lorsque le système est respectivement sous l’influence d’une perturbation déterministe haute fréquence ou d’une source de bruit. / This manuscript presents research aiming to show possible positive effects of deterministic and stochastic perturbations on the responses of different nonlinear systems. To that end, both numerical and experimental studies were carried out on two kinds of structures : an elementary electronic FitzHugh-Nagumo oscillator and an electrical line developed by resistively coupling 45 elementary cells. In the first section, the elementary cell characterization was undertaken in a deterministic regime. In the presence of a bichromatic stimulus, it is shown that when the low frequency component is subthreshold, its detection can be maximized for an optimal magnitude of the second component thanks to vibrational resonance. Next, it is established that this resonance may be enhanced for specific frequencies of the second component ; this phenomenon is referred to as frequency resonance. Furthermore, white and colored noise sources effects on vibrational resonance are reported. Then, for any other bichromatic excitation configuration, attention was focused on ghost stochastic resonance. Contrary to the other phenomena introduced in this manuscript, this one differs in the fact that the frequency of interest in the system output is here not applied on the input. Finally, the last part of the manuscript is devoted to the study of the coupled structure. It is shown that information propagation through line cells can be enhanced by vibrational propagation and noise assisted propagation phenomena. These nonlinear effects respectively occur when the system is under a high frequency deterministic perturbation or a random noise source.

<p>Doktorska disertacija je posvećena primeni teorije polugrupa operatora na re&scaron;avanje dve klase Cauchy-jevih početnih problema. U prvom delu smo<br />ispitivali parabolične stohastičke parcijalne diferencijalne jednačine (SPDJ-ne), odredjene sa dva tipa operatora: linearnim zatvorenim operatorom koji<br />generi&scaron;e <em>C</em>₀&minus;polugrupu i linearnim ograničenim operatorom kombinovanim<br />sa Wick-ovim proizvodom. Svi stohastički procesi su dati Wiener-It&ocirc;-ovom<br />haos ekspanzijom. Dokazali smo postojanje i jedinstvenost re&scaron;enja ove klase<br />SPDJ-na. Posebno, posmatrali smo i stacionarni slučaj kada je izvod po<br />vremenu jednak nuli. U drugom delu smo konstruisali kompleksne stepene<br /><em>C</em>-sektorijalnih operatora na sekvencijalno kompletnim lokalno konveksnim<br />prostorima. Kompleksne stepene operatora smo posmatrali kao integralne<br />generatore uniformno ograničenih analitičkih <em>C</em>-regularizovanih rezolventnih<br />familija, i upotrebili dobijene rezultate na izučavanje nepotpunih Cauchy-jevih problema vi&scaron;3eg ili necelog reda.</p> / <p>The doctoral dissertation is devoted to applications of the theory<br />of semigroups of operators on two classes of Cauchy problems. In the first<br />part, we studied parabolic stochastic partial differential equations (SPDEs),<br />driven by two types of operators: one linear closed operator generating a<br /><em>C</em>₀&minus;semigroup and one linear bounded operator with Wick-type multipli-cation. All stochastic processes are considered in the setting of Wiener-It&ocirc;<br />chaos expansions. We proved existence and uniqueness of solutions for this<br />class of SPDEs. In particular, we also treated the stationary case when the<br />time-derivative is equal to zero. In the second part, we constructed com-plex powers of <em>C</em>&minus;sectorial operators in the setting of sequentially complete<br />locally convex spaces. We considered these complex powers as the integral<br />generators of equicontinuous analytic <em>C</em>&minus;regularized resolvent families, and<br />incorporated the obtained results in the study of incomplete higher or frac-tional order Cauchy problems.</p>

En aquest treball s'ha estudiat la dinàmica dels làsers de classe A i de classe B en termes del potencial de Lyapunov. En el cas que s'injecti un senyal al làser o es modulin alguns dels paràmetres, apareix un comportament moltmés complex i s'estudia el conjunt de bifurcacions.1) Als làsers de classe A, la dinàmica determinista s'ha interpretat com el moviment damunt el potencial de Lyapunov. En la dinàmica estocàstica s'obté un flux sostingut per renou per a la fase del camp elèctric.2) Per als làsers de classe A amb senyal injectat, s'ha descrit el conjunt de bifurcacions complet i s'ha determinat el conjunt d'amplituds i freqüències en el quals el làser responajustant la seva freqüència a la del camp extern. 3) S'ha obtingut un potencial de Lyapunov pels làsers de classe B, només vàlid en el cas determinista, que inclou els termes de saturació de guany i d'emissió espontània.4) S'ha realitzat un estudi del conjunt de bifurcacions parcial al voltant del règim tipus II de la singularitat Hopf--sella--node en un làser de classe B amb senyal injectat.5) S'han identificat les respostes òptimes pels làsers de semiconductor sotmesos a modulació periòdica externa. S'han obtingut les corbes que donen la resposta màxima per cada tipus de resonància en el pla definit per l'amplitud relativa de modulació i la freqüència de modulació. / In this work we have studied the dynamics of both class A and class B lasers in terms of Lyapunov potentials. In the case of an injected signal or when some laser parameters are modulated, and more complex behaviour is expected, the bifurcation set is studied. The main results are the following:1) For class A lasers, the deterministic dynamics has been interpreted as a movement on the potential landscape. In the stochastic dynamics we have found a noise sustained flow for the phase of the electric field. 2) For class A lasers with an injected signal, we have been able to describe the whole bifurcation set of this system and to determine the set of amplitudes frequencies for which the laser responds adjusting its frequency to that of the external field. 3) In the case of class B lasers, we have obtained a Lyapunov potential only valid in the deterministic case, including spontaneous emission and gain saturation terms. The fixed point corresponding to the laser in the on state has been interpreted as a minimum in this potential. Relaxation to this minimum is reached through damped oscillations. 4) We have performed a study of the partial bifurcation set around the type II regime of the Hopf-saddle-node singularity in a class B laser with injected signal. 5) We have identified the optimal responses of a semiconductor laser subjected to an external periodic modulation. The lines that give a maximum response for each type of resonance are obtained in the plane defined by the relative amplitude modulation and frequency modulation.