[pt] O objetivo deste trabalho é desenvolver uma solução completa para o traçado de raios de cenas dinâmicas utilizando a GPU. Para que este algoritmo atinja desempenho interativo, é necessário utilizar uma estrutura espacial para reduzir os testes de interseção entre raios e triângulos da cena. Observa-se que, quando há movimento na cena, é necessário atualizar esta estrutura de aceleração, seja alterando-a parcialmente ou reconstruindo-a inteiramente. Adotamos a segunda estratégia por ser capaz de tratar o caso geral de movimento não-estruturado. Como a construção da estrutura deve ser feita da forma mais eficiente possível, escolhemos utilizar uma Grade Uniforme como foco de nossa pesquisa. Suas vantagens incluem um algoritmo de construção simples e um percurso de raios eficiente. Para explorar o poder de processamento em paralelo de uma GPU, é necessário manter os dados da cena e da estrutura de aceleração dentro da placa gráfica, evitando transferências custosas de memória entre a GPU e a CPU. Propomos neste trabalho uma técnica para construir uma grade uniforme inteiramente na GPU. Usando nosso método, é possível reconstruir toda a estrutura em poucos milissegundos, enquanto mantém-se a alta qualidade da grade obtida. Além disso, propomos uma implementaçoes do algoritmo de traçado de raios de forma a aproveitar o processamento em paralelo da GPU. Nosso procedimento é implementado inteiramente dentro da placa gráfica, onde há acesso direto para os dados dos triângulos da cena, bem como as informações da grade uniforme construída. Utilizando a solução proposta, somos capazes de obter taxas de visualização interativas mesmo para cenas com movimentos não-estruturados, incluindo texturas, sombras e até mesmo reflexões. / [en] We present a technique for ray tracing dynamic scenes using the GPU. In order to achieve interactive rendering rates, it is necessary to use a spatial structure to reduce the number of ray-triangle intersections performed. Whenever there is movement in the scene, this structure is entirely rebuilt. This way, it is possible to handle general unstructured motion. For this purpose, we have developed an algorithm for reconstructing Uniform Grids entirely inside the graphics hardware. In addition, we present ray-traversal and shading algorithms fully implemented on the GPU, including textures, shadows and reections. Combining these techniques, we demonstrate interactive ray tracing performance for dynamic scenes, even with unstructured motion and advanced shading effects.

Les travaux présentés dans ce mémoire portent sur l’étude et le développement d’une méthode de mesure d’épaisseur et de reconstruction 3D d’objets en verre creux, dans le cadre d’une application de contrôle industriel sur ligne de production. Les techniques de mesure d’épaisseur de parois transparentes sont actuellement très limitées par leur faible tolérance angulaire sur l’orientation des surfaces. Elles nécessitent une rotation de l’article devant le capteur, donc une manutention contraignante. De plus, les surfaces considérées étant spéculaires, les techniques de numérisation standard ne fonctionnent pas. Nous avons développé deux méthodes basées sur l’observation des réflexions d’une source diffuse ponctuelle sur les deux surfaces de la paroi de l’objet. Une observation simple à l’aide d’une caméra implique une ambiguïté entre l’orientation et la position de chacune des deux surfaces. Notre première méthode consiste à utiliser des informations de polarisation pour lever cette ambiguïté. Des résultats expérimentaux, obtenus avec un ou plusieurs points source témoignent de la faisabilité de la mesure. La seconde méthode consiste à utiliser une seconde caméra, associée à un dispositif optique spécifique. Ceci permet de trianguler directement les points d’incidence sur les deux surfaces. De plus, la source peut être étendue verticalement, de manière à augmenter la taille de la zone inspectée. Les résultats obtenus avec cette seconde technique attestent de la possibilité d’une mise en oeuvre industrielle. / The work presented in this thesis focuses on the study and development of a method for measuringthe thickness and reconstructing the 3D shape of glass containers, within the frame of an industrialon-line quality monitoring. Measuring thickness techniques of transparent objects are currently limited by their lack of toleranceregarding the surface orientation at a given reflection point. A rotation of the article in front ofthe sensor is thus required, which adds complexity to the mechanical handling of the item beingmeasured. Moreover, conventional scanning techniques do not work well on specular surfaces.We developed two methods based on the observation of the reflections by both surfaces of the wallof a given item of a diffuse, point light source. Imaging these reflections with a plain camera leavesan ambiguity between the orientation and position of each surface. Our first approach uses light polarization to remove this ambiguity. Experimental results, obtainedwith one or several point sources have demonstrated the feasibility of the measure. The second approach uses a second camera, associated with a specific optical design, whichenables a mere triangulation of the incident points on the two surfaces. Moreover, the light sourcecan be vertically extended to increase the size of the inspected area. The results obtained with thissecond technique show that industrial implementation is feasible.

Milieu où les ultrasons se propagent environ 10% plus rapidement qu’au sein des tissus environnants, le cristallin est connu pour être la source majeure d’aberrations de phase du milieu oculaire. De fait, l’échographie ophtalmique trans cornéenne est affectée par ses effets qui se manifestent sur les B-scans par une dégradation marquée de la résolution spatiale et du contraste, accompagnée de plus, d’importantes distorsions, particulièrement notables au niveau du fond de l’œil. Face à ce problème et en vue de l’arrivée prochaine de barrettes US HF dans la pratique ophtalmologique, un beamforming adapté a été développé au cours de cette thèse. Basé sur un lancer de rayon à 2 points fixes, il permet le calcul de délais de focalisation qui compensent les aberrations induites par le cristallin, en prenant en compte les effets réfractifs à son interface avec les humeurs. Les résultats obtenus in vitro et ex vivo avec une barrette 20MHz et un échographe de recherche (ECODERM) sont rapportés. / In ophthalmic ultrasonography the crystalline lens is known to be the main source of phase aberration, as ultrasounds (US) propagate about 10% faster than in the surrounding intra-ocular medium. Hence, it impairs significantly both spatial and contrast resolution of axial B-scans, and in addition causes important distortion, especially on the ocular fundus. To deal with this issue and in view of the next coming of US arrays in ophthalmologic practice, we developed in this thesis an adapted beamforming (BF) free from crystalline lens aberrations. It lies on a two point ray tracing approach to compute focusing delays that take into account crystalline lens aberrations including refraction at the interface. Initially developed considering a uniform US velocity in the lens, the adapted BF has been extended to consider the velocity gradient that exists in the real lens. In vitro and ex vivo results obtained with a 20 MHz linear array driven by a US research scanner (the ECODERM) are reported.

In 3D modelling software, deformations are used to add, to remove, or to modify geometric features of existing 3D models to create new models with similar but slightly different details. Traditional techniques for deforming virtual 3D models require users to explicitly define control points and regions of interest (ROIs), and to define precisely how to deform ROIs using control points. The awkwardness of defining these factors in traditional 3D modelling software makes it difficult for people with limited experience of 3D modelling to deform existing 3D models as they expect. As applications which require virtual 3D model processing become more and more widespread, it becomes increasingly desirable to lower the "difficulty of use" threshold of 3D model deformations for users. This thesis argues that the user experience, in terms of intuitiveness and ease of use, of a user interface for deforming virtual 3D models, can be greatly enhanced by employing sketch-based 3D model deformation techniques, which require the minimal quantities of interactions, while keeping the plausibility of the results of deformations as well as the responsiveness of the algorithms, based on modern home grade computing devices. A prototype system for sketch-based 3D model deformations is developed and implemented to support this hypothesis, which allows the user to perform a deformation using a single deforming stroke, eliminating the need to explicitly select control points, the ROI and the deforming operation. GPU based accelerations have been employed to optimise the runtime performance of the system, so that the system is responsive enough for real-time interactions. The studies of the runtime performance and the usability of the prototype system are conducted to provide evidence to support the hypothesis.

The master’s thesis deals with photometry units used in light engineering. There are defined the methods of modelling the illumination systems, their advantages, disadvantages and possibilities of using. Furthermore, the thesis deals with modelling of two illumination systems by ray-tracing in programming environment of MATLAB. The master’s thesis describes the analysis of the problem and the program designing process. Acquired model results are then compared with the real measurements of both illumination systems which allow the discussion of the results and the deviations of the models.

This thesis is aimed at simulation of auditory space. It describes the most frequently used panning method: Vector Base Amplitude Panning. Also is focused on image source method, which allows computing the parameters of direct sound wave and reflections in rectangular room. This method is compared with ray–tracing method, which is also often used. It deals with the matter of the frequency – dependent absorption of materials in reflection of the sound wave against the wall. On the basis of these information two applications were designed in MATLAB development environment. The first one allows the simulation of auditory space. The other one is the application for sound source panning by its azimuth and distance.

La classification des cibles sous-marines est principalement basée sur l'analyse de l'ombre acoustique. La nouvelle génération des sonars d'imagerie fournit une description plus précise de la rétrodiffusion de l'onde acoustique par les cibles. Par conséquent, la combinaison de l'analyse de l'ombre et de l'écho est une voie prometteuse pour améliorer la classification automatique des cibles. Quelques systèmes performants de classification automatique des cibles s'appuient sur un modèle pour faire l'apprentissage au lieu d'utiliser uniquement des réponses expérimentales ou simulées de cibles pour entraîner le classificateur. Avec une approche basée modèle, un bon niveau de performance en classification peut être obtenu si la modélisation de la réponse acoustique de la cible est suffisamment précise. La mise en œuvre de la méthode de classification a nécessité de modéliser avec précision la réponse acoustique des cibles. Le résultat de cette modélisation est un simulateur d'images sonar (SIS). Comme les sonars d'imagerie fonctionnent à haute et très haute fréquence le modèle est basé sur le lancer de rayons acoustiques. Plusieurs phénomènes sont pris en compte pour augmenter le réalisme de la réponse acoustique (les effets des trajets multiples, l'interaction avec le fond marin, la diffraction, etc.). La première phase du classificateur utilise une approche basée sur un modèle. L'information utile dans la signature acoustique de la cible est nommée « A-scan ». Dans la pratique, l'A-scan de la cible détectée est comparé à un ensemble d'A-scans générés par SIS dans les mêmes conditions opérationnelles. Ces gabarits (A-scans) sont créés en modélisant des objets manufacturés de formes simples et complexes (mines ou non mines). Cette phase intègre un module de filtrage adapté pour permettre un résultat de classification plus souple capable de fournir un degré d'appartenance en fonction du maximum de corrélation obtenu. Avec cette approche, l'ensemble d'apprentissage peut être enrichi afin d'améliorer la classification lorsque les classes sont fortement corrélées. Si la différence entre les coefficients de corrélation de l'ensemble de classes les plus probables n'est pas suffisante, le résultat est considéré ambigu. Une deuxième phase est proposée afin de distinguer ces classes en ajoutant de nouveaux descripteurs et/ou en ajoutant davantage d'A-scans dans la base d'apprentissage et ce, dans de nouvelles configurations proches des configurations ambiguës. Ce processus de classification est principalement évalué sur des données simulées et sur un jeu limité de données réelles. L'utilisation de l'A-scan a permis d'atteindre des bonnes performances de classification en mono-vue et a amélioré le résultat de classification pour certaines ambiguïtés récurrentes avec des méthodes basées uniquement sur l'analyse d'ombre. / Underwater target classification is mainly based on the analysis of the acoustic shadows. The new generation of imaging sonar provides a more accurate description of the acoustic wave scattered by the targets. Therefore, combining the analysis of shadows and echoes is a promising way to improve automated target classification. Some reliable schemes for automated target classification rely on model based learning instead of only using experimental samples of target acoustic response to train the classifier. With this approach, a good performance level in classification can be obtained if the modeling of the target acoustic response is accurate enough. The implementation of the classification method first consists in precisely modeling the acoustic response of the targets. The result of the modeling process is a simulator called SIS (Sonar Image Simulator). As imaging sonars operate at high or very high frequency the core of the model is based on acoustical ray-tracing. Several phenomena have been considered to increase the realism of the acoustic response (multi-path propagation, interaction with the surrounding seabed, edge diffraction, etc.). The first step of the classifier consists of a model-based approach. The classification method uses the highlight information of the acoustic signature of the target called « A-scan ». This method consists in comparing the A-scan of the detected target with a set of simulated A-scans generated by SIS in the same operational conditions. To train the classifier, a Template base (A-scans) is created by modeling manmade objects of simple and complex shapes (Mine Like Objects or not). It is based on matched filtering in order to allow more flexible result by introducing a degree of match related to the maximum of correlation coefficient. With this approach the training set can be extended increasingly to improve classification when classes are strongly correlated. If the difference between the correlation coefficients of the most likely classes is not sufficient the result is considered ambiguous. A second stage is proposed in order to discriminate these classes by adding new features and/or extending the initial training data set by including more A-scans in new configurations derived from the ambiguous ones. This classification process is mainly assessed on simulated side scan sonar data but also on a limited data set of real data. The use of A-scans have achieved good classification performances in a mono-view configuration and can improve the result of classification for some remaining confusions using methods only based on shadow analysis.

In our work an optical bus-coupler is proposed, which enables easy bidirectional connection between two waveguides without interrupting the bus using a core-to-core coupling principle. With bended waveguides the coupling ratio can be tuned by adjusting the overlap area of the two cores. In order to ensure large overlap areas at short coupling lengths, the waveguides have rectangular cross sections. To examine the feasibility of this coupling concept a simulation was performed, which is presented in this paper. Due to multimode waveguides, used in short range data communication, a non-sequential ray tracing simulation is reasonable. Simulations revealed that the bending of the waveguide causes a redistribution of the energy within the core. Small radii push the main energy to the outer region of the core increasing the coupling efficiency. On the other hand, at excessive lowered bend radii additional losses occur (due to a coupling into the cladding), which is why an optimum has to be found. Based on the simulation results it is possible to derive requirements and design rules for the coupling element.

This work makes basic analysis of historical and current algorithms for smoke animation. Modern approaches to rendering volumetric data are briefly described. We choose algorithms for implementation on basis of this analysis. These algorithms are described in detail and we make emphasis on their important properties according to dedication of this work. Detailed description of implementation follows along with performance measurement. Conclusion evaluates results of work and proposes possible extensions.

This thesis presents research towards enabling micromachined millimeter and submillimeter wave focal plane arrays (FPAs). The FPAs operate under the following principle: a switch network consisting of microelectromechanical (MEMS) switches, integrated with micromachined waveguides, is used to feed an array of antenna elements, located in the focal plane of a high-gain quasi-optical system. Hence, it is possible to switch between a set of narrow beams in different directions. Such beam steering systems are needed for future millimeter and submillimeter wave imaging and communication systems. The contributions to future MEMS-switchable FPAs presented here are organized in three papers, as described below. Paper I presents a criterion on the spacing between adjacent FPA elements which results in -3 dB overlap between the switched beams, for the special case when an extended hemispherical dielectric lens is used as the optical system. A key step towards this criterion is a closed-form relation between the scan angle and the FPA element's position, which results in an expression for the effective focal length of extended hemispherical lenses. A comparison with full-wave simulations demonstrates an excellent agreement with the presented theoretical results. Finally, it is shown that the maximum feasible FPA spacing when using an extended hemispherical lens is about 0.7 wavelengths. Paper II presents a numerical study of silicon-micromachined planar extended hemispherical lenses, with up to three matching regions used to reduce internal reflections. The effective permittivity of the matching regions is tailor-made by etching periodic holes in the silicon wafer. The optimal thickness and permittivity of the matching regions were determined using TRF optimization, in order to yield the maximum wide-band aperture efficiency and small side-lobes. We introduce a new matching region geometry, referred to as shifted-type matching regions, and it is demonstrated that using three shifted-type matching regions results in twice as large aperture efficiency as compared to using three conventional concentric-type matching regions. Paper III presents a submillimeter-wave single-pole single-throw (SPST) 500-750 GHz MEMS waveguide switch, based on a MEMS-reconfigurable surface inserted between two waveguide flanges. A detailed design parameter study is carried out to select the best combination of the number of horizontal bars and vertical columns of the MEMS-reconfigurable surface, for achieving a low insertion loss in the transmissive state and a high isolation in the blocking state. A method is presented to model the non-ideal electrical contacts between the vertical cantilevers of the MEMS surface, with an excellent agreement between the simulated and measured isolation. It is shown that the isolation can be improved by replacing an ohmic contact by a new, capacitive contact. The measured isolation of the switch prototype is better than 19 dB and the measured insertion loss is between 2.5 and 3 dB. / Denna avhandling presenterar forskning som syftar till att möjliggöra fokalplans-gruppantenner (FPAs) för våglängder i millimeter och submillimeterområdet. Principen för en sådan FPAs funktion är följande: ett nätverk bestående av mikroelektromekaniska (MEMS) switchar, används för att välja mellan de olika antenn-elementen i en gruppantenn, som placerats i fokalplanet av ett optiskt system. Därmed blir det möjligt att välja från en uppsättning av smala lober i olika riktningar. Sådana lob-styrningssystem behövs för framtida radar- och kommunikationssystem i millimeter och submillimeterområdet. Resultaten är uppdelade i tre vetenskapliga artiklar, som beskrivs nedan. I den första artikeln (Paper I) presenteras ett villkor för avståndet mellan närliggande FPA-element som resulterar i -3 dB överlappning mellan de switchade loberna, för specialfallet då en förlängd hemisfärisk lins används som optiskt system. Det viktigaste steget mot att hitta detta villkor är att bestämma en analytisk relation mellan avsökningsvinkeln och FPA-elementens position. Detta resulterar i ett uttryck för den effektiva fokallängden för denna typ av lins. En utmärkt överensstämelse har funnits mellan dessa relationer och simuleringar. Slutligen visas det att de största möjliga FPA-avstånden för en förlängd hemisfärisk lins är ungefär 0.7 våglängder, vilket uppnås för linser med låg permittivitet. I den andra artikeln (Paper II) presenteras en numerisk studie av plana förlängda hemisfäriska linser, som kan produceras från en kiselskiva. Linserna har upp till tre matchningsregioner, som används för att reducera interna reflektioner. Den effektiva permittiviteten av de matchande regionerna skräddarsys genom etsning av periodiska hål i kiselskivan. Den optimala tjockleken och permittiviteten av de matchande regionerna har bestämts med hjälp av TRF-optimering, för att ge maximal bredbandig direktivitet och minimala sidlober. En ny geometri introduceras för matchningsregionerna, som vi kallar matchningsregioner av skiftad typ. Vi visar att användning av tre matchningsregioner av skiftad typ resulterar i en dubbelt så hög apertur-effektivitet, jämfört med att använda tre konventionella matchningsregioner av koncentrisk typ. I den tredje artikeln (Paper III) presenteras en MEMS-switch för rektangulära vågledare, för frekvensområdet 500-750 GHz. Baserat på en designparameterstudie har den bästa kombinationen av antalet horisontella rader och vertikala kolumner hos den MEMS-konfigurerbara ytan valts ut, för att uppnå låga förluster i det öppna tillståndet och hög isolation i det blockerande tillståndet. I artikeln presenteras en metod för att modellera icke-perfekta elektriska kontakter mellan de fixerade och de rörliga delarna i MEMS-ytan. Denna metod uppvisar en utmärkt överensstämmelse mellan den simulerade och den uppmätta isolationen. Vi visara att isolationen kan förbättras med hjälp av en ny typ av kapacitiv kontakt. Den uppmätta isolationen hos den presenterade switch-prototypen är högre än 19 dB, och den uppmätta förlusten är mellan 2.5 och 3 dB. / <p>QC 20161206</p>