Interpolation is essential in digital image processing, especially magnification. Many different approaches to interpolation specific to magnification have been developed in an effort to overcome the shortcomings of bilinear and bicubic interpolation. One of these approaches, Constraint-Based Interpolation, produces an image that is free of jaggies and has less blurring than bilinear or bicubic interpolation. Although Constraint-Based Interpolation produces a visually pleasing image, there are user-chosen parameters that make the algorithm difficult to use. In this thesis we propose a method for automatic selection of those parameters and an extension of Constraint-Based Interpolation to other forms of image manipulation, such as skew, rotation, warp, or any other invertable image transformation. By extending Constaint-Based Interpolation the same improvements observed in magnification could be observed in these other image transformations.

No description available.

Segmented analysis of subcortical structures within the nonhuman primate can potentially have a profound impact on studying the relationship between volumetric characteristics and alcohol dependencies. Image segmentations have been widely used in quantifying structural information. There are a variety of methods in which users can extract desired structures from a medical image; ranging from manual segmentations to fully-automated segmentations and 2-D to 3-D. The implications of this possibility can have tremendous applicability to medical research and diagnosis. The primary goal of my thesis is to investigate different implementation methodologies for segmenting subcortical structures such as the hippocampus and striatum and then apply that knowledge towards the development of an approach to segment these two structures from a group of alcohol-dependent Rhesus Macaque monkeys. Using the Level Set Deformable Model (LSDM) with a priori structural information, a series of T1-weighted MR images of Rhesus Macaque hippocampi and striatum were segmented in an effort to compare the structural hippocampal and striatal volumes between early and late stages of alcohol dependency. The results suggest that the volumes of both subcortical structures are affected negatively by alcoholism. Volume deficits of as much as 5% for the hippocampus and 8% for the caudate were found. / Master of Science

La macroségrégation est un défaut connu dans les procédés de coulées industrielles. La genèse de ce défaut est la conséquence de l'interaction complexe entre la microségrégation ou la distribution des espèces chimiques à l'échelle de la microstructure et les mouvements des phases liquide et solides. Les hétérogéneités de concentration en solutés à l'échelle de la pièce peuvent être rédhibitoires vis-à-vis de la qualité du produit.Dans ce travail, on propose un modèle numérique pour simuler et prédire la formation des macroségrégations en coeur des pièces d'alliages multi-constitués, induites par des variations thermiques et solutales dans la phase liquide. Dans un premier temps, on considère que le métal solidifie à volume constant. Dans ce contexte, la convection thermosolutale est étudiée ainsi que son influence sur la formation des canaux ségrégées à différentes échelles de modélisation. Dans un deuxième temps, le modèle vise à prédire les macroségrégations en présence de changement de volume du métal, dont la cause principale est le retrait à la solidification, pouvant être à l'origine du phénomène de ségrégation inverse. La surface entre le métal et le gaz environnant au cours du retrait évolue pendant le retrait en fonction du chemin de solidification qui varie avec la macroségrégation. Cette évolution d'interface est suivi par la méthode level set. Des prédictions de concentration moyenne, couplées aux bases de données thermodynamiques pour mieux prédire les chemins de solidification des alliages multi-constitués, sont analysées et comparées avec des résultats expérimentaux. Finalement, des calculs de solidification en microgravité sont présentées, simulant un essai expérimental dans le contexte du projet CCEMLCC lancé par l'Agence Spatiale Européenne. Les résultats en fin de solidification montrent un accord acceptable quant à la forme et l'élongation des échantillons solidifiés. Ces calculs sont faits avec des approximations binaire, ternaire et quaternaire d'une même nuance d'acier utilisée dans les essais en microgravité. / Macrosegregation is key defect in industrial casting processes. During solidification, solute redistribution at the scale of microstructure, also known as microsegregation, take place with complex interactions, in order to form one or more solid phases. These interactions between microsegregation and movements of liquid and solid phases may lead to macrosegregations. These solute heterogeneities spanning on a larger scale, may result in a bad casting quality. In this thesis, we propose a numerical model to simulate and predict macrosegregations occurring in the centre of multicomponent alloys, caused by thermal and solutal variations in the liquid phase. First, we assume that the metallic alloy solidifies with a constant volume. In this context, we study the influence of thermosolutal convection on the formation of channel segregations, at different modelling scales. The second part of this modelling work consider solidification while the metallic alloy's volume is decreasing, mainly due to overall density variation, also known as solidification shrinkage, possibly leading to the so-called inverse segregation phenomenon, appearing on the alloy's skin. In the context of solidification shrinkage, the shape of the metal's boundary with surrounding gases varies according to a constantly changing solidification path due to macrosegregation. The level set method is therefore used to track its evolution with time. Composition predictions, coupled with thermodynamic database mappings for more accurate multicomponent solidification paths, are analysed and compared to existing experimental setups. Finally, simulations of a reduced-gravity solidification cases are performed, mocking an experimental benchmark from the CCEMLCC project launched by the European Space Agency. The results after complete solidification show acceptable agreement for the final shape, compared to experimental results. These computations were performed with binary, ternary and quaternary approximations of the same steel grade which was used in reduced-gravity experiments.

A otimização topológica de estruturas está relacionada à concepção de projetos que executem suas funções com nível de segurança adequado empregando a quantidade mínima de material. Neste trabalho, determina-se a geometria ótima de estruturas planas por meio do acoplamento do Método dos Elementos de Contorno (MEC) ao Método Level Set (MLS). O algoritmo é composto por 3 etapas: problema mecânico, otimização topológica e reconstrução da estrutura. O problema mecânico é resolvido pelas equações algébricas do MEC. A otimização topológica é determinada pelo MLS, este representa a geometria do corpo e suas evoluções por meio da função Level Set (LS) avaliada em seu nível zero. Na reconstrução realiza-se o remalhamento, pois a cada iteração a estrutura é modificada. O acoplamento proposto resulta na geometria ótima da estrutura sem a necessidade da aplicação de filtros. Os exemplos analisados mostram que algoritmo desenvolvido capta adequadamente a geometria ótima das estruturas. Com esse trabalho, avança-se no campo das aplicações do acoplamento MEC-MLS e no desenvolvimento de soluções inovadoras para problemas complexos de engenharia. / In general, the topological optimization of structures is related to design projects that perform their functions with appropriate security levels using the minimum amount of material. This research determines the optimal geometry of 2D structures by coupling the Boundary Blement Method (BEM) to Level Set Method (LSM). The algorithm consists of 3 steps: mechanical model, topology optimization and structure reconstruction. The mechanical model is solved by BEM algebraic equations. The topology optimization is determined using the MLS, the geometry of the body is determined by the Level Set (LS) function evaluated at the zero level. The reconstruction achieves the remeshing, because for each iteration of the structure is modified. The proposed coupling results in the optimal geometry of the structure without the filters application. The examples show that the algorithm developed captures adequately the optimal geometry of the structures. With this dissertation, it is possible advance in the field of applications of the BEM - LSM and develop innovative solutions to complex engineering problems.

A otimização topológica de estruturas está relacionada à concepção de projetos que executem suas funções com nível de segurança adequado empregando a quantidade mínima de material. Neste trabalho, determina-se a geometria ótima de estruturas planas por meio do acoplamento do Método dos Elementos de Contorno (MEC) ao Método Level Set (MLS). O algoritmo é composto por 3 etapas: problema mecânico, otimização topológica e reconstrução da estrutura. O problema mecânico é resolvido pelas equações algébricas do MEC. A otimização topológica é determinada pelo MLS, este representa a geometria do corpo e suas evoluções por meio da função Level Set (LS) avaliada em seu nível zero. Na reconstrução realiza-se o remalhamento, pois a cada iteração a estrutura é modificada. O acoplamento proposto resulta na geometria ótima da estrutura sem a necessidade da aplicação de filtros. Os exemplos analisados mostram que algoritmo desenvolvido capta adequadamente a geometria ótima das estruturas. Com esse trabalho, avança-se no campo das aplicações do acoplamento MEC-MLS e no desenvolvimento de soluções inovadoras para problemas complexos de engenharia. / In general, the topological optimization of structures is related to design projects that perform their functions with appropriate security levels using the minimum amount of material. This research determines the optimal geometry of 2D structures by coupling the Boundary Blement Method (BEM) to Level Set Method (LSM). The algorithm consists of 3 steps: mechanical model, topology optimization and structure reconstruction. The mechanical model is solved by BEM algebraic equations. The topology optimization is determined using the MLS, the geometry of the body is determined by the Level Set (LS) function evaluated at the zero level. The reconstruction achieves the remeshing, because for each iteration of the structure is modified. The proposed coupling results in the optimal geometry of the structure without the filters application. The examples show that the algorithm developed captures adequately the optimal geometry of the structures. With this dissertation, it is possible advance in the field of applications of the BEM - LSM and develop innovative solutions to complex engineering problems.

Dans ce travail, nous nous sommes intéressés à la simulation numérique de suspensions denses d'objets immergés dans un fluide. En s'inspirant d'une méthode de segmentation d'image, nous avons développé une méthode efficace de capture d'interface permettant d'une part de localiser les structures immergées et d'autre part de gérer les contacts numériques entre les structures.Le domaine fluide/structure est représenté à l'aide de trois fonctions labels et deux fonctions distances qui permettent de localiser chaque structure et son plus proche voisin.Les interfaces sont capturées par une seule fonction level set, celle-ci est ensuite transportée par la vitesse du fluide ou par la vitesse de chaque structure. Un algorithme de multi-label fast marching permet de réinitialiser à chaque pas de temps les fonctions labels et distances dans un périmètre proche des interfaces.La gestion des contacts numériques est effectuée grâce à une force répulsive à courte portée prenant en compte l'interaction entre les objets les plus proches.Dans un premier temps, la méthode est appliquée à l'évolution de solides rigides immergés.Un modèle de pénalisation global couplé aux fonctions labels permet de calculer en une seule fois l'ensemble des vitesses des structures rigides. Les résultats obtenus montrent l'efficacité de la méthode à gérer un grand nombre de solides.Nous avons ensuite appliqué la méthode des suspensions de vésicules immergées. Ce type de simulation requiert le calcul des forces élastiques et de courbures exercées sur les membranes. Grâce au modèle proposé, seulement une force élastique et une force de courbure sont calculées pour l'ensemble des membranes à l'aide de la fonction level set et des fonctions labels. / In this work, we propose a method to efficiently capture an arbitrary number of fluid/solid or fluid/fluid interfaces, in a level-set framework. This technique, borrowed from image analysis, is introduced in the context of the interaction of several bodies immersed in a fluid. A configuration of the bodies in the fluid/structure domain is described by three label maps providing the first and second neighbours, and their associated distance functions. Only one level set function captures the union of all interfaces and is transported with the fluid velocity or with a global velocity field which takes into account the velocity of each structure. A multi-label fast marching method is then performed in a narrow-band around the interfaces allowing to update the label and distance functions. Within this framework, the numerical treatment of contacts between the structures is achieved by a short-range repulsive force depending on the distance between the closest bodies.The method is validated through the simulation of a dense suspension of rigid bodies immersed in an incompressible fluid. A global penalization model uses the label maps to follow the solid bodies altogether without a separate computation of each body velocity. Consequently, the method shows its efficiency when dealing with a large number of rigid bodies. We also investigate the numerical simulation of vesicle suspensions for which a computation of elastic and bending forces on membranes is required. In the present model, only one elastic and bending force is computed for the whole set of membranes according to the level set function and the label maps.

Le procédé de fabrication par RTM (Resin Transfer Molding) a été étudié numériquement à l’aide de la méthode XFEM (eXtended Finite Element Method) combinée avec la méthode Level set. La méthode XFEM permet d’obtenir une bonne précision numérique de la pression près du front d’écoulement, où son gradient est discontinu. Les fonctions de forme enrichies de la méthode XFEM sont proposées à l’aide des valeurs de Level set en vue de décrire correctement l’interpolation avec le front d’écoulement. En plus, la méthode de Level set est utilisée pour transporter le front d’écoulement à chaque pas de temps durant le remplissage du moule. Les valeurs de Level set sont calculées à l’aide d’une méthode de Galerkin implicite. Le solveur multi-frontal d’IPSAP a été utilisé pour la résolution du système. Cette étude a été validée en comparaison avec les solutions analytiques.En outre, une méthode de localisation avec XFEM et la méthode Level set a été proposée afin d’améliorer l’efficacité de calcul. Elle permet de réduire le domaine de calcul près du front d’écoulement. Par conséquent, le temps de calcul est fortement réduit grâce à cette méthode. Un test d’efficacité a été fait avec des modèles simples en écoulement laminaire ou radial.Quelques exemples d’application sont présentés pour illustrer la capacité de cette méthode. Une pale d’éolienne a également traitée comme application industrielle. Enfin, une interface d’utilisateur graphique a été développée en vue de fournir une facilité des pré- et post-processus. / Numerical simulation for Resin Transfer Molding (RTM) manufacturing process is attempted by using the eXtended Finite Element Method (XFEM) combined with the level set method. XFEM allows to obtaining a good numerical precision of the pressure near the resin flow front, where its gradient is discontinuous. The enriched shape functions of XFEM are derived by using the level set values so as to correctly describe the interpolation with the resin flow front. In addition, the level set method is used to transport the resin flow front at each time step during the mold filling. The level set values are calculated by an implicit characteristic Galerkin FEM. The multi-frontal solver of IPSAP is adopted to solve the system. This work is validated by comparing the obtained results with analytic solutions.Moreover, a localization method of XFEM and level set method is proposed to increase the computing efficiency. The computation domain is reduced to the small region near the resin flow front. Therefore, the total computing time is strongly reduced by it. The efficiency test is made with simple channel or radial flow models. Several application examples are analyzed to demonstrate ability of this method. A wind turbine blade is also treated as industrial application. Finally, a Graphic User Interface (GUI) tool is developed so as to make easy the pre/post-processing of the simulation.

La prédiction de l’état de fissuration est un enjeu crucial pour l’analyse des structures en béton armé, qui nécessite le recours à la modélisation et à la simulation numérique. Le calcul par éléments finis des structures en béton armé pose au moins deux problèmes majeurs :d’une part il existe peu de modèles permettant de traiter à la fois l’initiation, la propagation et l’ouverture des fissures, d’autre part le diamètre généralement faible des armatures métalliques par rapport aux dimensions des structures étudiées nécessite des maillages particulièrement fins. On propose donc des solutions à ces deux problématiques basées sur l’utilisation de fonctions de niveau (level set). L’endommagement et la fissuration du béton sont modélisés à l’aide de l’approche TLS (Thick Level Set). Cette méthode,développée en tant que méthode de régularisation des modèles d’endommagement locaux, utilise une level set afin d’introduire une longueur caractéristique. Cela permet de rendre aisée la localisation de la position des fissures, et donc d’enrichir le champ de déplacement parla méthode des éléments finis étendus (X-FEM) afin de modéliser l’ouverture des macro-fissures. Concernant la modélisation des armatures, une nouvelle approche multidimensionnelle est proposée. Une représentation volumique des armatures par la méthode X-FEM est utilisée dans les zones d’intérêt afin d’obtenir des résultats précis tout en simplifiant la procédure de maillage, tandis qu’une représentation linéique est utilisée dans le reste de la structure afin de réduire le nombre de degrés de liberté du calcul. La méthode de transition développée ici permet d’assurer la cohérence des résultats obtenus / Prediction of cracking is a key point for the analysis ofreinforced concrete structures, which requires the use of Modeling and numerical simulation. The analysis of reinforced concrete structures using the finite element method raises two issues: on one hand, few models areable to deal with the initiation, the propagation and the opening of cracks, on the other hand the diameter of thereinforcements which is usually small compared to the dimensions of the structures necessitates very fine meshes. Some solutions to these two problematics areproposed, based on the use of level set functions.Damage and cracking of concrete are modeled using theThick Level Set (TLS) approach. This method,developped as a mean to regularize local damagemodels, uses a level set to introduce a characteristic length. It makes the location of the cracks easy, whichallows to enrich the displacement field with the eXtendedFinite Element Method (X-FEM) in order to model the macro-cracks opening. Concerning the modeling of thereinforcements, a new multidimensionnal approach isproposed. A volumic representation of the reinforcements with the X-FEM method is used in the zones of interest to get accurate results while simplifying the meshing process, whereas a lineic representation isused elsewhere to decrease the number of degrees of freedom. The developed transition method insures the consistency of the results.

Motivé par l'étude de tissus biologiques, ce travail contribue aux développements d'outils numériques permettant de prédire la réponse mécanique de matériaux hétérogènes non linéaires dans lesquels les énergies d'interfaces deviennent prépondérantes. Ainsi, une méthode d'homogénéisation multi échelle combinée à une technique de réduction de modèle basée sur la décomposition orthogonale aux valeurs propres est proposée dans un cadre thermique et hyperélastique. Les énergies d'interfaces entre les différentes phases des composites sont décrites par un modèle d'interface cohérent et prises en compte numériquement par une approche liant la méthode des éléments finis étendus et la méthode level-set. Une étude de l'étalement d'une cellule vivante entre deux lamelles fixes est ensuite réalisée. Les deux modèles utilisés pour les simulations montrent que l'assemblage cortex d'actine-membrane plasmique ne joue qu'un rôle minime dans la réponse mécanique cellulaire / Motivated by the study of biological tissues, this work contributes to developing numerical tools to predict the mechanical response of nonlinear heterogeneous materials in which the energies of interfaces can no longer be ignored. First, a computational homogenization strategy combined with a model reduction technique based on the proper orthogonal decomposition is implemented in the cases of large elastic deformations and highly nonlinear conduction. The interfaces between the different phases of a composite are described by means of a coherent interface model and taken into account numerically by an extended finite element method in tandem with a level-set technique. Finally, experimental results of single cell spreading between two fixed parallel microplates are exploited through finite element modelling. Our two models show that the bilayer membrane and the actin cortex do not play a significant role in the cell mechanical response