La production d'aubes en TA6V par matriçage à chaud (vers 940°C) donne occasionnellement l i eu à l'apparition de défauts de surface. La lubrification est assurée côté outils par un lubrifiant graphité, côté pièces par un film d'émail, mélange d'oxydes métalliques et de liant s organiques déposé à froid sur l'ébauche et d'épaisseur de l'ordre de 100 à 200 μm. Après une étude bibliographique, nous étudions l'évolution du film d'émail lors du préchauffage sur échantillons et au cours du procédé industriel sur pièces (aubes, " soucoupes "). Le préchauffage confère au film d'émail des propriétés proches de celles des verres métalliques, réduit fortement son épaisseur , mais laisse une porosité et une rugosité variables qui ne permettent pas la mesure de sa viscosité par des méthodes classiques. Le défaut est une accumulation locale d'émail et de graphite et résulte d'une conjugaison de facteurs : excès d'épaisseur et de porosité résiduelle du film d'émail , température d'outil trop élevée, avarie du film lubrifiant... Un mécanisme de formation du défaut par écoulement de l'émail à l'interface métal-outil du centre vers le bord du contact et rupture de la croûte du film à l'extérieur du contact est proposé. Pour déterminer la viscosité des films d'émail, nous développons deux essais de déformation isotherme de métal revêtu d'émail : indentation à basse température fournissant la température de transition fragile ductile correspondant à 101 0 Pas environ ; compression à température élevée permettant d'estimer la viscosité par mesure de l'épaisseur du film piégée. La mise au point de cet essai s'est faite sur un verre étalon par modification des géométries (piégeage entre outil et lopin, puis entre deux lopins) et par simulation numérique du piégeage (Forge2® multi-mat.) afin de corriger l'analyse théorique de Wilson. Enfin. un essai anisotherme d'écrasement de cylindre sur génératrice permet d'étudier l'effet du lubrifiant, de l'épaisseur d'émail et de l a température d'outil sur le frottement et la répartition d'émail à l'interface. La simulations numérique (Forge3®) permet d'identifier le frottement et d'analyser l'évolution d'épaisseur de l'émail : elle dépend de l'étirement superficiel du métal et des gradients de température superficielle et de pression de contact. L'écoulement de l'émail est mis en équation et sa résolut ion à partir des données issues de la simulation du forgeage (pression de contact, scission de frottement, température superficielle) permettrait d'évaluer la redistribution d'émail au cours du matriçage et les risques d'apparition du défaut.

3D representation nowadays has attracted much more public attention than ever before. One of the most important techniques in this field is depth extraction. In this thesis, we first introduce a well-known stereo matching method using color segmentation and belief propagation, and make an implementation of this framework. The color-segmentation based stereo matching method performs well recently, since this method can keep the object boundaries accurate, which is very important to depth map. Based on the implemented framework of segmentation based stereo matching, we proposed a color segmentation based 2D-to-3D video conversion method using high quality motion information. In our proposed scheme, the original depth map is generated from motion parallax by optical flow calculation. After that we employ color segmentation and plane estimation to optimize the original depth map to get an improved depth map with sharp object boundaries. We also make some adjustments for optical flow calculation to improve its efficiency and accuracy. By using the motion vectors extracted from compressed video as initial values for optical flow calculation, the calculated motion vectors are more accurate within a shorter time compared with the same process without initial values. The experimental results shows that our proposed method indeed gives much more accurate depth maps with high quality edge information. Optical flow with initial values provides good original depth map, and color segmentation with plane estimation further improves the depth map by sharpening its boundaries.

La compréhension des mécanismes de transition La compréhension des mécanismes de transition d'écoulements décollés laminaires trouve un intérêt particulier de par sa présence au sein de nombreuses applications industrielles. Notamment, la prédiction précise de la première bifurcation associée à un tel écoulement et la construction de modèles réduits, pourrait autoriser des développements d'outils de contrôle efficaces, permettant une augmentation des performances aérodynamiques des profils considérés. Dans ce cadre, nous proposons à travers ce mémoire une analyse de stabilité originale, prenant en compte le caractère fortement non parallèle de l'écoulement. En particulier, une étude de stabilité linéaire "globale", sur une configuration académique d'écoulement décollé de plaque plane, révèle l'influence des modes globaux bidimensionnels et tridimensionnels, dans la dynamique en espace et en temps de la perturbation, en fonction du type de forçage appliqué. En outre, une telle étude identifie des mécanismes non détectés par les approches classiques, comme une très large amplification transitoire 2D de la perturbation le long de la couche de mélange, et un mécanisme centrifuge pouvant conduire à une tridimensionnalisation du décollement. Enfin, la décomposition de la perturbation dans la base de modes globaux s'avère une réduction de modèle pertinente, dans l'objectif d'un contrôle efficace.

The Brunswick No. 6 area, which is located in the Bathurst Mining Camp, New Brunswick, Canada, is the focus of this thesis. Almost a decade ago, in order to improve the understanding of the crustal structures and explore for new mineral deposits at depth, three 2D seismic profiles totaling about 30 km and 3D seismic data covering an area of about 38 km2 were acquired from the study area. Petrophysical properties including compressional-wave velocity and density were also measured in two deep boreholes in the area. These data were recovered and reanalyzed, and the improved seismic images interpreted as the main part of this PhD thesis. A prestack DMO and poststack migration algorithm was considered for processing both 2D and 3D data. Processing of 2D data revealed shallow and deep reflections, which correlate well with surface geology. Steeply-dipping reflections, some of which could host mineral deposits, were imaged down to a depth of 6-7 km. Processing of 3D data showed similar results to the processed 2D profiles. Nevertheless, the non-orthogonal nature of the 3D survey, combined with irregular distribution of offsets, azimuths and trace midpoints, caused a severe acquisition footprint masking reflections in the DMO-corrected unmigrated stacked cube. An FK-dip filter in the wavenumber domain was designed to reduce the effects of the acquisition footprint. To better understand wave propagation and scattering effects, calculated acoustic impedance log from the available borehole data was used to estimate vertical scale length using a von Karman autocorrelation function. 2D synthetic models representative of heterogeneity in the area were generated accounting for the estimated scale length. Numerical modeling was used to study the scattering effects on the synthetic models, where some predefined targets were superimposed in the provided 2D heterogeneous medium. The effects of variable source frequency, longer horizontal scale length and petrophysical fluctuations of heterogeneous medium were also investigated. The modeling results indicate that, in the presence of large horizontal, but small vertical scale lengths (structural anisotropy), the identification of mineral deposits is possible in the unmigrated stacked sections, but can be challenging in the migrated sections.

3D representation nowadays has attracted much more public attention than ever before. One of the most important techniques in this field is depth extraction. In this thesis, we first introduce a well-known stereo matching method using color segmentation and belief propagation, and make an implementation of this framework. The color-segmentation based stereo matching method performs well recently, since this method can keep the object boundaries accurate, which is very important to depth map. Based on the implemented framework of segmentation based stereo matching, we proposed a color segmentation based 2D-to-3D video conversion method using high quality motion information. In our proposed scheme, the original depth map is generated from motion parallax by optical flow calculation. After that we employ color segmentation and plane estimation to optimize the original depth map to get an improved depth map with sharp object boundaries. We also make some adjustments for optical flow calculation to improve its efficiency and accuracy. By using the motion vectors extracted from compressed video as initial values for optical flow calculation, the calculated motion vectors are more accurate within a shorter time compared with the same process without initial values. The experimental results shows that our proposed method indeed gives much more accurate depth maps with high quality edge information. Optical flow with initial values provides good original depth map, and color segmentation with plane estimation further improves the depth map by sharpening its boundaries.

This dissertation describes the design and implementation of a system that has been designed to extract 3D information from pairs of 2D images. System input consists of two images taken by an ordinary digital camera. System output is a full 3D model extracted from 2D images. There are no assumptions about the positions of the cameras during the time when the images are being taken, but the scene must not undergo any modifications. The process of extracting 3D information from 2D images consists of three basic steps. First, point matching is performed. The main contribution of this step is the introduction of an approach to matching image segments in the context of an approximation space. The second step copes with the problem of estimating external camera parameters. The proposed solution to this problem uses 3D geometry rather than the fundamental matrix widely used in 2D to 3D conversion. In the proposed approach (DirectGS), the distances between reprojected rays for all image points are minimised. The contribution of the approach considered in this step is a definition of an optimal search space for solving the 2D to 3D conversion problem and introduction of an efficient algorithm that minimises reprojection error. In the third step, the problem of dense matching is considered. The contribution of this step is the introduction of a proposed approach to dense matching of 3D object structures that utilises the presence of points on lines in 3D space. The theory and experiments developed for this dissertation demonstrate the usefulness of the proposed system in the process of digitizing 3D information. The main advantage of the proposed approach is its low cost, simplicity in use for an untrained user and the high precision of reconstructed objects. / October 2007

This dissertation describes the design and implementation of a system that has been designed to extract 3D information from pairs of 2D images. System input consists of two images taken by an ordinary digital camera. System output is a full 3D model extracted from 2D images. There are no assumptions about the positions of the cameras during the time when the images are being taken, but the scene must not undergo any modifications. The process of extracting 3D information from 2D images consists of three basic steps. First, point matching is performed. The main contribution of this step is the introduction of an approach to matching image segments in the context of an approximation space. The second step copes with the problem of estimating external camera parameters. The proposed solution to this problem uses 3D geometry rather than the fundamental matrix widely used in 2D to 3D conversion. In the proposed approach (DirectGS), the distances between reprojected rays for all image points are minimised. The contribution of the approach considered in this step is a definition of an optimal search space for solving the 2D to 3D conversion problem and introduction of an efficient algorithm that minimises reprojection error. In the third step, the problem of dense matching is considered. The contribution of this step is the introduction of a proposed approach to dense matching of 3D object structures that utilises the presence of points on lines in 3D space. The theory and experiments developed for this dissertation demonstrate the usefulness of the proposed system in the process of digitizing 3D information. The main advantage of the proposed approach is its low cost, simplicity in use for an untrained user and the high precision of reconstructed objects.

This dissertation describes the design and implementation of a system that has been designed to extract 3D information from pairs of 2D images. System input consists of two images taken by an ordinary digital camera. System output is a full 3D model extracted from 2D images. There are no assumptions about the positions of the cameras during the time when the images are being taken, but the scene must not undergo any modifications. The process of extracting 3D information from 2D images consists of three basic steps. First, point matching is performed. The main contribution of this step is the introduction of an approach to matching image segments in the context of an approximation space. The second step copes with the problem of estimating external camera parameters. The proposed solution to this problem uses 3D geometry rather than the fundamental matrix widely used in 2D to 3D conversion. In the proposed approach (DirectGS), the distances between reprojected rays for all image points are minimised. The contribution of the approach considered in this step is a definition of an optimal search space for solving the 2D to 3D conversion problem and introduction of an efficient algorithm that minimises reprojection error. In the third step, the problem of dense matching is considered. The contribution of this step is the introduction of a proposed approach to dense matching of 3D object structures that utilises the presence of points on lines in 3D space. The theory and experiments developed for this dissertation demonstrate the usefulness of the proposed system in the process of digitizing 3D information. The main advantage of the proposed approach is its low cost, simplicity in use for an untrained user and the high precision of reconstructed objects.

<p>Detta är en reflexiv rapport om arbetsmetoden bakom skapandet av en filmtrailer som ska marknadsföra den fiktiva filmen om Gabriel Glömmer baserad på en bok av Ulf Löfgren. Boken handlar om en tillbakadragen pojke med en livlig fantasi som drömmer sig till en sagovärld där han får uppleva äventyr och möta nya vänner. Texten har ett fokus på skapandet av miljöer i 3D, med mål att behålla den grafiska stilen från 2D illustrationer som finns i boken. Resultatet av projektet är en filmtrailer med längden 1 minut och 40 sekunder. Filmtrailerns bilder har en uppbyggnad som försöker efterlikna de illustrationer som finns i boken.</p>

<p>Rapporten tar upp mina reflektioner dels angående projektarbetet både på ett personligt plan samt samarbetet mellan oss två som gjorde projektet. Vad har fungerat och vad har fungerat mindre bra under projekttiden? Från 2D till 3D är en röd tråd genom hela rapporten då vi gått från 2D-referernsbilder i barnboken Gabriel Glömmer till utrenderade 3D-sekvenser. Jag går igenom vad jag gjort och vad andra gjort under projekttiden då vi haft hjälp av lite olika människor. Dessa människor presenteras snabbt, något annat som presenteras är boken och min relation till densamma. Resultatet av projektarbetet blev att vi lyckades med de uppställde mål vi hade innan projektet startades.</p>