Dissertations / Theses on the topic 'Direction Reconstruction'

Neutrinos are the perfect cosmic messengers when it comes to investigating the most violent and mysterious astronomical and cosmological events in the Universe. The interaction probability of neutrinos is small, and the flux of high-energy neutrinos decreases quickly with increasing energy. In order to find high-energy neutrinos, large bodies of matter needs to be instrumented. A proposed detector station design called ARIANNA is designed to detect neutrino interactions in the Antarctic ice by measuring radio waves that are created due to the Askaryan effect. In this paper, we present a method based on state-of-the-art machine learning techniques to reconstruct the direction of the incoming neutrino, based on the radio emission that it produces. We trained a neural network with simulated data, created with the NuRadioMC framework, and optimized it to make the best possible predictions. The number of training events used was on the order of 106. Using two different emission models, we found that the network was able to learn and generalize on the neutrino events with good precision, resulting in a resolution of 4-5°. The model could also make good predictions on a dataset even if it was trained with another emission model. The results produced are promising, especially due to the fact that classical techniques have not been able to reproduce the same results without having prior knowledge of where the neutrino interaction took place. The developed neural network can also be used to assess the performance of other proposed detector designs, to quickly and reliably give an indication of which design might yield the most amount of value to the scientific community.
Neutriner är de perfekta kosmiska budbärarna när det kommer till att undersöka de mest våldsamma och mystiska astronomiska och kosmologiska händelserna i vårt universum. Sannolikheten för en neutrinointeraktion är dock liten, och flödet av högenergetiska neutriner minskar kraftigt med energin. För att hitta dessa högenergetiska neutriner måste stora volymer av materia instrumenteras. Ett förslag på en design för en detektorstation kallas ARIANNA, och är framtagen för att detektera neutrinointeraktioner i den antarktiska isen genom att mäta radiopulser som bildas på grund av Askaryan-effekten. I denna rapport presenterar vi en metod baserad på toppmoderna maskininlärningstekniker för att rekonstruera riktningen på en inkommande neutrino, utifrån den radiostrålning som produceras. Vi tränade ett neuralt nätverk med simulerade data, som skapades med hjälp av ramverket NuRadioMC, och optimerade nätverket för att göra så bra förutsägelser som möjligt. Antalet interaktionshändelser som användes för att träna nätverket var i storleksordningen 106. Genom att undersöka två olika emissionsmodeller fann vi att nätverket kunde generalisera med god precision. Detta resulterade i en upplösning på 4-5°. Modellen kunde även göra goda förutsägelser på en datamängd trots att nätverket var tränat med en annan emissionsmodell. De resultat som metoden framtog är lovande, särskilt med avseende på att tidigare klassiska metoder inte har lyckats reproducera samma resultat utan att metoden redan innan vet var i isen som neutrinointeraktionen skedde. Nätverket kan också komma att användas för att utvärdera prestandan hos andra designförslag på detektorstationer för att snabbt och säkert ge en indikation på vilken design som kan tillhandahålla mest vetenskapligt värde.

L'intérêt de l'échantillonnage compressé dans l'imagerie ultrasonore a été récemment évalué largement par plusieurs équipes de recherche. Suite aux différentes configurations d'application, il a été démontré que les données RF peuvent être reconstituées à partir d'un faible nombre de mesures et / ou en utilisant un nombre réduit d'émission d'impulsions ultrasonores. Selon le modèle de l'échantillonnage compressé, la résolution des images ultrasonores reconstruites à partir des mesures compressées dépend principalement de trois aspects: la configuration d'acquisition, c.à.d. l'incohérence de la matrice d'échantillonnage, la régularisation de l'image, c.à.d. l'a priori de parcimonie et la technique d'optimisation. Nous nous sommes concentrés principalement sur les deux derniers aspects dans cette thèse. Néanmoins, la résolution spatiale d'image RF, le contraste et le rapport signal sur bruit dépendent de la bande passante limitée du transducteur d'imagerie et du phénomène physique lié à la propagation des ondes ultrasonores. Pour surmonter ces limitations, plusieurs techniques de traitement d'image en fonction de déconvolution ont été proposées pour améliorer les images ultrasonores. Dans cette thèse, nous proposons d'abord un nouveau cadre de travail pour l'imagerie ultrasonore, nommé déconvolution compressée, pour combiner l'échantillonnage compressé et la déconvolution. Exploitant une formulation unifiée du modèle d'acquisition directe, combinant des projections aléatoires et une convolution 2D avec une réponse impulsionnelle spatialement invariante, l'avantage de ce cadre de travail est la réduction du volume de données et l'amélioration de la qualité de l'image. Une méthode d'optimisation basée sur l'algorithme des directions alternées est ensuite proposée pour inverser le modèle linéaire, en incluant deux termes de régularisation exprimant la parcimonie des images RF dans une base donnée et l'hypothèse statistique gaussienne généralisée sur les fonctions de réflectivité des tissus. Nous améliorons les résultats ensuite par la méthode basée sur l'algorithme des directions simultanées. Les deux algorithmes sont évalués sur des données simulées et des données in vivo. Avec les techniques de régularisation, une nouvelle approche basée sur la minimisation alternée est finalement développée pour estimer conjointement les fonctions de réflectivité des tissus et la réponse impulsionnelle. Une investigation préliminaire est effectuée sur des données simulées
The interest of compressive sampling in ultrasound imaging has been recently extensively evaluated by several research teams. Following the different application setups, it has been shown that the RF data may be reconstructed from a small number of measurements and/or using a reduced number of ultrasound pulse emissions. According to the model of compressive sampling, the resolution of reconstructed ultrasound images from compressed measurements mainly depends on three aspects: the acquisition setup, i.e. the incoherence of the sampling matrix, the image regularization, i.e. the sparsity prior, and the optimization technique. We mainly focused on the last two aspects in this thesis. Nevertheless, RF image spatial resolution, contrast and signal to noise ratio are affected by the limited bandwidth of the imaging transducer and the physical phenomenon related to Ultrasound wave propagation. To overcome these limitations, several deconvolution-based image processing techniques have been proposed to enhance the ultrasound images. In this thesis, we first propose a novel framework for Ultrasound imaging, named compressive deconvolution, to combine the compressive sampling and deconvolution. Exploiting an unified formulation of the direct acquisition model, combining random projections and 2D convolution with a spatially invariant point spread function, the benefit of this framework is the joint data volume reduction and image quality improvement. An optimization method based on the Alternating Direction Method of Multipliers is then proposed to invert the linear model, including two regularization terms expressing the sparsity of the RF images in a given basis and the generalized Gaussian statistical assumption on tissue reflectivity functions. It is improved afterwards by the method based on the Simultaneous Direction Method of Multipliers. Both algorithms are evaluated on simulated and in vivo data. With regularization techniques, a novel approach based on Alternating Minimization is finally developed to jointly estimate the tissue reflectivity function and the point spread function. A preliminary investigation is made on simulated data

Electrical Impedance Tomography is an imaging technique that aims to reconstruct the inner conductivity distribution of a medium starting from a set of measured voltages registered by a series of electrodes that are positioned on the surface of the medium. Such technique was used for the first time in geological studies in 1930 and then applied to industrial procedures. The first clinical use of EIT dates back to 1987. In 2018 EIT was validated in tissue engineering as a noninvasive and label-free imaging and monitoring tool for cell distribution (cell growth, differentiation and tissue formation) in 3D scaffolds. EIT problem can be split into a Forward and an Inverse problem. The aim of Forward EIT is to define the set of measured voltages starting from a known conductivity distribution. If the forward problem is characterized by a nonlinear mapping, called Forward Operator, from the conductivity distribution to the measured voltages, inverse EIT consists of inverting the Forward Operator. This leads to an ill-posed problem which requires regularization, either in the model or in the numerical method that is applied to define the solution. The inverse problem is modelled as a Nonlinear Least Squares problem, where one seeks to minimize the mismatch beetween the measured voltages and the ones generated by the reconstructed conductivity. Reconstruction techniques require the introduction of a regularization term which forces the reconstructed data to stick to certain prior information. In this dissertation, some state-of-the-art regularization methods are analyzed and compared via EIDORS, a specific software for EIT problems. The aim is to reconstruct the variation in conductivity within a 2D section of a 3D scaffold. Furthermore a variational formulation on a 2D mesh for a space-variant regularization is proposed, based on a combination of high order and nonconvex operators, which respectively seek to recover piecewise inhomogeneous and piecewise linear regions.

2013/2014
Per uno studio completo che parte dalla materia oscura e va all'origine e propagazione dei raggi cosmici, quello multi canale è uno degli approcci migliori per risolvere i quesiti irrisolti della fisica delle astroparticelle. GAMMA-400, grazie alla sua natura duale, dedita allo studio di raggi cosmici (elettroni fino alle energie del TeV e protoni e nuclei fino a 10^{15}-10^{16} eV) e raggi gamma (da 50 MeV fino a qualche TeV), si dedicherà allo studio di questi problemi. Lo scopo di questa tesi è lo studio delle prestazioni di GAMMA-400 per l'osservazione dei raggi gamma. Due diverse configurazioni della geometria sono state studiate: la "baseline" e la cosiddetta configurazione "enhanced". Le principali differenze tra queste due configurazioni si trovano nel tracciatore e nel calorimetro. Il tracciatore della "baseline" è composto da dieci piani di silicio, otto dei quali comprendono anche uno strato di ~0.1 X_0 di tungsteno. Il tracciatore della configurazione "enhanced" è invece composto da 25 piani di silicio inframezzati da uno strato di tungsteno di ~0.03 X_0. Il calorimetro della "baseline" è diviso in due sezioni: una parte composta da due piani di ioduro di cesio e silicio (chiamata "pre-shower") e una seconda parte composta da 28x28x12 cubi di ioduro di cesio. Il calorimetro della configurazione "enhanced" è invece composto solo da 20x20x20 cubi di ioduro di cesio. Per stimare le prestazioni ho sviluppato un algoritmo di ricostruzione della direzione del raggio gamma incidente. La ricostruzione può fare uso delle informazioni provenienti dal tracciatore, dal "pre-shower" o dal calorimetro, sia combinandole che singolarmente. Le direzioni ottenuta grazie alle informazioni del solo "pre-shower" o del solo calorimetro, anche se di minor risoluzione, possono essere utili per aumentare il numero di fotoni visti ad alta energia e per fornire le informazioni necessarie all'osservazione di transienti con i telescopi Cherenkov a terra. La risoluzione angolare utilizzando il tracciatore è migliore nel caso della configurazione "enhanced". A basse energie ciò è possibile grazie al minore tungsteno, e di conseguenza minor "scattering" multiplo, presente all'interno del tracciatore. Il calorimetro più piccolo, e più profondo, seppur ostacolando la ricostruzione dell'energia di fotoni ad alta energia, produce anche un numero minore di particelle di "backsplash" che peggiorano la ricostruzione delle tracce. L'area efficace totale della "baseline", potendo contare su un calorimetro più grande ed il "pre-shower", è più grande rispetto alla configurazione "enhanced". La risoluzione angolare, l'area efficace e la strategia di osservazione dello strumento contribuiscono alla sensitività per sorgenti puntiformi. La sensitività totale dello strumento è migliore per la "baseline" per energie maggiori di 5 GeV. Ho implementato un set prelminare di condizioni di "trigger" per lo studio dei raggi gamma tramite l'utilizzo delle informazioni del tracciatore. La necessità di rigettare la maggior parte delle particelle cariche deriva dall'elevato fondo presente in orbita (~10^6 protoni per raggio gamma) e una capacità di "downlink" limitata (~100 GB/day). Tra le due configurazioni si nota una differenza di meno dell'1% nel numero rimanente di protoni. Seppur promettente, tale risultato deve essere migliorato e possibili miglioramenti sono descritti nella tesi. Gli algoritmi di ricostruzione e "trigger" sono applicati all'analisi della possibilità di studiare "gamma-ray burst" (GRB) con la principale strumentazione a bordo di GAMMA-400. Una stima del numero di eventi non ricostruiti, perchè avvengono nel tempo morto tra due "trigger", è effettuata tramite la simulazione di un ipotetico GRB accoppiata ai tempi di arrivo dei fotoni presi dai dati reali di due GRB osservati da Fermi. In nessuna delle due configurazioni è visibile una percentuale significativa di "pile-up". Anche aumentando il flusso dei GRB la percentuale di eventi non ricostruiti non supera mai il 6%. Nonostante questo risultato, molto dipenderà dal disegno finale dell’elettronica di lettura dei rivelatori che potrebbe aumentare i tempi morti dello strumento.
XXVII Ciclo
1987

La modélisation géométrique d'objets fabriqués par l'homme à partir de données 3D est l'un des plus grands défis de la vision par ordinateur et de l'infographie. L'objectif à long terme est de générer des modèles de type CAO de la manière la plus automatique possible. Pour atteindre cet objectif, des problèmes difficiles doivent être résolus, notamment (i) le passage à l'échelle du processus de modélisation sur des données d'entrée massives, (ii) la robustesse de la méthodologie contre des mesures d'entrées erronés, et (iii) la qualité géométrique des modèles de sortie. Les méthodes existantes fonctionnent efficacement pour reconstruire la surface des objets de forme libre. Cependant, dans le cas d'objets fabriqués par l'homme, il est difficile d'obtenir des résultats dont la qualité approche celle des représentations hautement structurées, comme les modèles CAO. Dans cette thèse, nous présentons une série de contributions dans ce domaine. Tout d'abord, nous proposons une méthode de classification basée sur l'apprentissage en profondeur pour distinguer des objets dans des environnements complexes à partir de nuages de points 3D. Deuxièmement, nous proposons un algorithme pour détecter des primitives planaires dans des données 3D à différents niveaux d'abstraction. Enfin, nous proposons un mécanisme pour assembler des primitives planaires en maillages polygonaux compacts. Ces contributions sont complémentaires et peuvent être utilisées de manière séquentielle pour reconstruire des modèles de ville à différents niveaux de détail à partir de données 3D aéroportées. Nous illustrons la robustesse, le passage à l'échelle et l'efficacité de nos méthodes sur des données laser et multi-vues stéréo sur des scènes composées d'objets fabriqués par l'homme
Geometric modeling of man-made objects from 3D data is one of the biggest challenges in Computer Vision and Computer Graphics. The long term goal is to generate a CAD-style model in an as-automatic-as-possible way. To achieve this goal, difficult issues have to be addressed including (i) the scalability of the modeling process with respect to massive input data, (ii) the robustness of the methodology to various defect-laden input measurements, and (iii) the geometric quality of output models. Existing methods work well to recover the surface of free-form objects. However, in case of manmade objects, it is difficult to produce results that approach the quality of high-structured representations as CAD models.In this thesis, we present a series of contributions to the field. First, we propose a classification method based on deep learning to distinguish objects from raw 3D point cloud. Second, we propose an algorithm to detect planar primitives in 3D data at different level of abstraction. Finally, we propose a mechanism to assemble planar primitives into compact polygonal meshes. These contributions are complementary and can be used sequentially to reconstruct city models at various level-of-details from airborne 3D data. We illustrate the robustness, scalability and efficiency of our methods on both laser and multi-view stereo data composed of man-made objects

Omni-directional (O-D) infrared (IR) vision is an effective capability for mobile systems in robotics, due to its advantages: illumination invariance, wide field-of-view, ease of identifying heat-emitting objects, and long term tracking without interruption. Unfortunately, O-D IR sensors have low resolution, low frame rates, high cost, sensor noise, and an increase in tracking time. In order to overcome these disadvantages, we propose an autonomous system application in indoor scenarios including 1) Dynamic 3D Reconstruction (D3DR) of the target view in real time images, 2) Human Behavior-based Target Tracking from O-D thermal images, 3) Thermal Multisensor Fusion (TMF), and 4) Visual Perception for Social Cognition from the motion behavior of the human target.

This study identifies Late Quaternary glacial ice-flow direction in east-central Ellesmere Island, N.W.T., by the study of its composition and its relation to bedrock sources. The spatial distribution of five till types suggests a hypothesis which attempts to reconcile the Innuitian Ice Sheet model of Blake with the Franklin Ice Complex model of England. Glaciers draining from ice sheets in northern Ellesmere Island and Greenland filled Kane Basin and flowed southward toward Smith Sound. When this southward-flowing ice reached the bottleneck at Smith Sound (40 km wide), it overrode the adjacent landmasses of Pim Island and Cape Herschel and forced Buchanan Bay ice southward through the channel that is now occupied by Rice Strait. Southward-flowing ice continued to drain towards Baffin Bay, overriding the coastal areas of Nares Strait at Wade Point and Cape Isabella. The results presented here partially support the Innuitian ice sheet hypothesis, but do not negate the Franklin Ice Complex model. (Abstract shortened by UMI.)

Nous proposons dans cette thèse de développer une méthode de restauration aveugle d'images flouées et bruitées où aucune connaissance a priori n'est exigée. Ce manuscrit est composé de trois chapitres : le 1er chapitre est consacré aux travaux de l'état de l'art. Les approches d'optimisation pour la résolution du problème de restauration y sont d'abord discutées. Ensuite les principales méthodes de restauration, dites semi-aveugles car nécessitant un minimum de connaissance a priori sont analysées. Parmi ces méthodes, cinq sont retenues pour évaluation. Le 2ème chapitre est dédié à la comparaison des performances des méthodes retenues dans le chapitre précédent. Les principaux critères objectifs d'évaluation de la qualité des images restaurées sont présentés. Parmi ces critères, la norme L1 de l'erreur d'estimation est sélectionnée. L'étude comparative menée sur une banque d'images monochromes, dégradées artificiellement par deux fonctions floues de supports différents et trois niveaux de bruit a permis de mettre en évidence les deux méthodes les plus pertinentes. La première repose sur une approche alternée mono-échelle où la PSF et l'image sont estimées dans une seule étape. La seconde utilise une approche hybride multi-échelle qui consiste tout d'abord à estimer de manière alternée la PSF et une image latente, puis dans une étape suivante séquentielle, à restaurer l'image. Dans l'étude comparative conduite, l'avantage revient à cette dernière. Les performances de ces méthodes serviront de référence pour comparer ensuite la méthode développée. Le 3ème chapitre porte sur la méthode développée. Nous avons cherché à rendre aveugle l'approche hybride retenue dans le chapitre précédent tout en améliorant la qualité d'estimation de la PSF et de l'image restaurée. Les contributions ont porté sur plusieurs points. Une première série d'améliorations concerne la redéfinition des échelles, celle de l'initialisation de l'image latente à chaque niveau d'échelle, l'évolution des paramètres pour la sélection des contours pertinents servant de support à l'estimation de la PSF et enfin, la définition d'un critère d'arrêt aveugle. Une seconde série de contributions a porté sur l'estimation aveugle des deux paramètres de régularisation impliqués pour éviter d'avoir à les fixer empiriquement. Chaque paramètre est associé à une fonction coût distincte l'une pour l'estimation de la PSF et la seconde pour l'estimation d'une image latente. Dans l'étape séquentielle qui suit, nous avons cherché à affiner le support de la PSF estimée dans l'étape alternée, avant de l'exploiter dans le processus de restauration de l'image. A ce niveau, la seule connaissance a priori nécessaire est une borne supérieure du support de la PSF. Les différentes évaluations conduites sur des images monochromes et hyperspectrales dégradées artificiellement par plusieurs flous de type mouvement, de supports différents, montrent une nette amélioration de la qualité de restauration obtenue par l'approche développée par rapport aux deux meilleures approches de l'état de l'art retenues
We propose in this thesis manuscript to develop a blind restoration method of single component blurred and noisy images where no prior knowledge is required. This manuscript is composed of three chapters: the first chapter focuses on state-of-art works. The optimization approaches for resolving the restoration problem are discussed first. Then, the main methods of restoration, so-called semi-blind ones because requiring a minimum of a priori knowledge are analysed. Five of these methods are selected for evaluation. The second chapter is devoted to comparing the performance of the methods selected in the previous chapter. The main objective criteria for evaluating the quality of the restored images are presented. Of these criteria, the l1 norm for the estimation error is selected. The comparative study conducted on a database of monochromatic images, artificially degraded by two blurred functions with different support size and three levels of noise, revealed the most two relevant methods. The first one is based on a single-scale alternating approach where both the PSF and the image are estimated alternatively. The second one uses a multi-scale hybrid approach, which consists first of alternatingly estimating the PSF and a latent image, then in a sequential next step, restoring the image. In the comparative study performed, the benefit goes to the latter. The performance of both these methods will be used as references to then compare the newly designed method. The third chapter deals with the developed method. We have sought to make the hybrid approach retained in the previous chapter as blind as possible while improving the quality of estimation of both the PSF and the restored image. The contributions covers a number of points. A first series concerns the redefinition of the scales that of the initialization of the latent image at each scale level, the evolution of the parameters for the selection of the relevant contours supporting the estimation of the PSF and finally the definition of a blind stop criterion. A second series of contributions concentrates on the blind estimation of the two regularization parameters involved in order to avoid having to fix them empirically. Each parameter is associated with a separate cost function either for the PSF estimation or for the estimation of a latent image. In the sequential step that follows, we refine the estimation of the support of the PSF estimated in the previous alternated step, before exploiting it in the process of restoring the image. At this level, the only a priori knowledge necessary is a higher bound of the support of the PSF. The different evaluations performed on monochromatic and hyperspectral images artificially degraded by several motion-type blurs with different support sizes, show a clear improvement in the quality of restoration obtained by the newly designed method in comparison to the best two state-of-the-art methods retained

Wave tank tests facilitate the understanding of how complex sea conditions influence the dynamics of man-made structures. If a potential deployment location is known, site data can be used to improve the relevance and realism of the test conditions, thus helping de-risk device development. Generally this data is difficult to obtain and even if available is used simplistically due to established practices and limitations of test facilities. In this work four years of buoy data from the European Marine Energy Centre is characterised and simulated at the FloWave Ocean Energy Research Facility; a circular combined wave-current test tank. Particular emphasis is placed on the characterisation and validation processes, aiming to preserve spectral and directional complexity of the site, whilst proving that the defined representative conditions can be effectively created. When creating representative site-specific sea states, particular focus is given to the application of clustering algorithms, which enable the entire spectral (frequency or directional) form to be considered in the characterisation process. This enables the true complex nature of the site to be considered in the data reduction process. Prior to generating and measuring the resulting sea states, issues with scaling are explored, the facility itself is characterised, and emphasis is placed on developing measurement strategies for the validation of directional spectra. Wave gauge arrays are designed and used to characterise various elements of the FloWave tank, including reflections, spatio-temporal variability and wave shape. A new method for directional spectrum reconstruction (SPAIR) is also developed, enabling more effective measurement and validation of the resulting directional sea states. Through comparison with other characterisation methods, inherent method-induced trade-offs are understood, and it is found that there is no absolute favourable approach, necessitating an application specific procedure. Despite this, a useful set of 'generic' sea states are created for the simulation of both production and extreme conditions. For sea state measurement, the SPAIR method is proven to be significantly more effective than current approaches, reducing errors and introducing additional capability. This method is used in combination with a directional wave gauge array to effectively measure, correct, and validate the resulting directional wave conditions. It is also demonstrated that site-specific wave-current scenarios can be effectively re-created, thus demonstrating that truly complex ocean conditions can be simulated at FloWave. This ability, along with the considered characterisation approach used, means that representative site-specific sea states can be simulated with confidence, increasing the realism of the test environment and helping de-risk device development.

Magnetic resonance imaging (MRI) is an imaging technique to visualize internal structures of the body. Diffusion MRI is an MRI modality that measures overall diffusion effect of molecules in vivo and non-invasively. Diffusion tensor imaging (DTI) is an extended technique of diffusion MRI. The major application of DTI is to measure the location, orientation and anisotropy of fiber tracts in white matter. It enables non-invasive investigation of major neural pathways of human brain, namely tractography. As spatial resolution of MRI is limited, it is possible that there are multiple fiber bundles within the same voxel. However, diffusion tensor model is only capable of resolving a single direction. The goal of this dissertation is to investigate complex anatomical structures using high angular resolution diffusion imaging (HARDI) data without any assumption on the parameters. The dissertation starts with a study of the noise distribution of truncated MRI data. The noise is often not an issue in diffusion tensor model. However, in HARDI studies, with many more gradient directions being scanned, the number of repetitions of each gradient direction is often small to restrict total acquisition time, making signal-to-noise ratio (SNR) lower. Fitting complex diffusion models to data with reduced SNR is a major interest of this study. We focus on fitting diffusion models to data using maximum likelihood estimation (MLE) method, in which the noise distribution is used to maximize the likelihood. In addition to the parameters being estimated, we use likelihood values for model selection when multiple models are fit to the same data. The advantage of carrying out model selection after fitting the models is that both the quality of data and the quality of fitting results are taken into account. When it comes to tractography, we extend streamline method by using covariance of the estimated parameters to generate probabilistic tracts according to the uncertainty of local tract orientations.

The objective of this thesis is to provide a complete pipeline that achieves an accurate reconstruction of the white matter fascicles using clinical diffusion images characterized by a low angular resolution. This involves (i) a diffusion model inferred in each voxel from the diffusion images and (ii) a tractography algorithm fed with these local models to perform the actual reconstruction of fascicles. Our contribution in diffusion modeling is a new statistical distribution, the properties of which are extensively studied. We model the diffusion as a mixture of such distributions, for which we design a model selection tool that estimates the number of mixture components. We show that the model can be accurately estimated from single shell low angular resolution diffusion images and that it provides specific biomarkers for studying tumors. Our contribution in tractography is an algorithm that approximates the distribution of fascicles emanating from a seed voxel. We achieve that by means of a particle filter better adapted to multi-modal distributions than the traditional filters. To demonstrate the clinical applicability of our tools, we participated to all three editions of the MICCAI DTI Tractography challenge aiming at reconstructing the cortico-spinal tract from single-shell low angular and low spatial resolution diffusion images. Results show that our pipeline provides a reconstruction of the full extent of the CST.

La FWI (Full Waveform Inversion) est un problème d'optimisation sous contraintes dédié à l'estimation des paramètres constitutifs du sous-sol à partir de mesures parcimonieuses des champs d'ondes sismiques. La FWI est fondée sur des approches locales d'optimisation et sur un espace de recherche réduit obtenu par projection de variables. La non linéarité et le caractère mal posé de la FWI sont deux difficultés majeures. Une source de non linéarité est liée au repliement de la phase, qui conduit à un minimum local dès que le modèle initial n'est pas suffisamment précis. Le caractère mal posé résulte de l'éclairage incomplet du sous-sol depuis la surface, le bruit et les couplages inter-paramètres. L'objectif de cette thèse est de réduire ces deux pathologies par de nouvelles approches d'optimisation et de régularisation. J'améliore tout d'abord la méthode d'inversion par reconstruction des champs d'onde (WRI : Wavefield Reconstruction Inversion). WRI étend l'espace de recherche en calculant les champs d'onde avec une relaxation de l'équation d'onde afin d'ajuster les données avec des modèles imprécis avant d'estimer les paramètres en minimisant les erreurs générées par cette relaxation. Quand ces deux estimations sont effectuées de manière alternée, WRI décompose l'inversion non linéaire en deux sous-problèmes linéaires en vertu de la bilinéarité de l'équation d'onde. WRI a été implémentée avec une méthode de pénalité, nécessitant une adaptation du paramètre de pénalité lors des itérations. Je remédie à cela avec ADMM (Alternating-Direction Method of Multipliers), qui concilie l'extension de l'espace de recherche et la précision de la solution au point de convergence avec un paramètre de pénalité fixe grâce à la mise à jour itérative des multiplicateurs de Lagrange. Une seconde contribution est l'implémentation de contraintes de bornes et de régularisation par variation totale (TV) dans WRI. Suivant la méthode de Split Bregman, des variables auxiliaires permettent de découpler les termes impliquant des normes ℓ2 et ℓ1 et de traiter les seconds efficacement avec des opérateurs de proximité. Ensuite, j'ai combiné une régularisation de Tikhonov et de TV par convolution infimale pour prendre en compte les différentes propriétés statistiques du milieu (constantes par morceau et lisses). Ma thèse aborde ensuite des reconstructions multi-paramètres. Je montre dans un premier temps que la bilinéarité de l'équation d'onde est vérifiée pour les équations de l'elastodynamique. Ensuite, je traite le cas de milieux acoustique VTI où je reconstruis conjointement la vitesse verticale et epsilon pour un modèle synthétique représentatif d'un champ pétrolier en mer du Nord. Je m'intéresse ensuite à l'imagerie de l'atténuation qui est introduite en domaine harmonique sous forme d'une vitesse complexe. J'étends WRI à la reconstruction de paramètres complexes tout en développant une régularisation adaptable à la vitesse réelle et au facteur de qualité. Durant les premières itérations, les champs d'onde reconstruits sont précis uniquement au voisinage des récepteurs. Les imprécisions de la phase pourraient avoir un rôle préjudiciable sur la solution de l'inversion. Afin de réduire cette empreinte, j'estime les paramètres par "phase retrieval", un processus qui vise la reconstruction d'un signal complexe à partir de l'amplitude de sa mesure linéaire. Une fois un premier modèle obtenu, je réinjecte l'information de la phase pour converger vers la solution finale. Je montre la pertinence de cette stratégie lorsque le modèle initial est homogène. WRI a été initialement développée dans le domaine fréquentiel car la reconstruction des champs d'onde y est raisonnablement aisée avec des méthodes d'algèbre linéaire. En domaine temporel, une approche fondée sur un schéma explicite d'intégration temporelle a été proposée mais repose sur une linéarisation autour des sources supposées connues
Full Waveform Inversion (FWI) is a PDE-constrained optimization which reconstructs subsurface parameters from sparse measurements of seismic wavefields. FWI generally relies on local optimization techniques and a reduced-space approach where the wavefields are eliminated from the variables. In this setting, two bottlenecks of FWI are nonlinearity and ill-posedness. One source of nonlinearity is cycle skipping, which drives the inversion to spurious minima when the starting subsurface model is not kinematically accurate enough. Ill-posedness can result from incomplete subsurface illumination, noise and parameter cross-talks. This thesis aims to mitigate these pathologies with new optimization and regularization strategies. I first improve the wavefield reconstruction method (WRI). WRI extends the FWI search space by computing wavefields with a relaxation of the wave equation to match the data from inaccurate parameters. Then, the parameters are updated by minimizing wave equation errors with either alternating optimization or variable projection. In the former case, WRI breaks down FWI into to linear subproblems thanks to wave equation bilinearity. WRI was initially implemented with a penalty method, which requires a tedious adaptation of the penalty parameter in iterations. Here, I replace the penalty method by the alternating-direction method of multipliers (ADMM). I show with numerical examples how ADMM conciliates the search space extension and the accuracy of the solution at the convergence point with fixed penalty parameters thanks to the dual ascent update of the Lagrange multipliers. The second contribution is the implementation of bound constraints and non smooth Total Variation (TV) regularization in ADMM-based WRI. Following the Split Bregman method, suitable auxiliary variables allow for the de-coupling of the ℓ1 and ℓ2 subproblems, the former being solved efficiently with proximity operators. Then, I combine Tikhonov and TV regularizations by infimal convolution to account for the different statistical properties of the subsurface (smoothness and blockiness). At the next step, I show the ability of sparse promoting regularization in reconstruction the model when ultralong offset sparse fixed-spread acquisition such as those carried out with OBN are used. This thesis continues with the extension of the ADMM-based WRI to multiparameter reconstruction in vertical transversely isotropic (VTI) acoustic media. I first show that the bilinearity of the wave equation is satisfied for the elastodynamic equations. I discuss the joint reconstruction of the vertical wavespeed and epsilon in VTI media. Second, I develop ADMM-based WRI for attenuation imaging, where I update wavefield, squared-slowness, and attenuation in an alternating mode since viscoacoustic wave equation can be approximated, with a high degree of accuracy, as a multilinear equation. This alternating solving provides the necessary flexibility to taylor the regularization to each parameter class and invert large data sets. Then, I overcome some limitations of ADMM-based WRI when a crude initial model is used. In this case, the reconstructed wavefields are accurate only near the receivers. The inaccuracy of phase of the wavefields may be the leading factor which drives the inversion towards spurious minimizers. To mitigate the role of the phase during the early iterations, I update the parameters with phase retrieval, a process which reconstructs a signal from magnitude of linear mesurements. This approach combined with efficient regularizations leads to more accurate reconstruction of the shallow structure, which is decisive to drive ADMM-based WRI toward good solutions at higher frequencies. The last part of this PhD is devoted to time-domain WRI, where a challenge is to perform accurate wavefield reconstruction with acceptable computational cost

The accumulated importance of the patent law morality exclusion over the recent decades has been prompted by the increased protection sought for biotechnological inventions. This raises specific ethical concerns such as the respect for human dignity. The morality exclusion is regarded as an indispensable safeguard in the patent system of the 21st century, but its application and scope remains unclear, not least due to a division of competence between the European Union (EU), the European Patent Organisation and national Member States. In addition, the international Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) under the auspices of the World Trade Organization (WTO) sets minimum standards for the operation of the morality exclusion. In this study the scope, interpretation and application of the morality exclusion in European and international patent law is described and analysed by applying the legal dogmatic method. The object of study is the European patent law morality exclusion in terms of Article 53(a) and Rule 28 of the European Patent Convention and Article 6 of Directive 98/44/EC on the legal protection of biotechnological inventions. A connected purpose of this study is to describe, problematize and analyse the compliance by the regional rules (EU, European Patent Organisation) with the international framework provided by Article 27.2 of the TRIPS Agreement. Although this study is not limited to specific technologies, the majority of issues raised are done so in relation to the morality exclusion concerning biotechnological inventions, which is reflected in this work. Furthermore, the theoretical framework is dependent on the functioning of the morality exclusion in the respective systems, both regional (EU, European Patent Organisation) and international (WTO). This approach necessitates a consideration of the characteristics of each legal system. This thesis uncovers the differences in the interpretation of the morality exclusion (both within as well as between the legal systems), which causes legal uncertainty in this particularly complex field where patent law and ethics interact. The effects of such variations on the scope and application of the morality exclusion are clarified and critically reviewed. According to the wording of the morality exclusion, the commercial exploitation of an invention with regard to morality or ordre public is a central prerequisite in the assessment of excluded subject matter. Consequently, the interpretation of this particular requirement is subject to critical review, since its implications are decisive for the scope of application of the morality exclusion. It is furthermore questioned whether the European morality exclusions are TRIPS-compliant, and the possible effects (and available remedies) of non-compliance for the EU legal order are analysed. This in turn involves issues of norm conflicts in international and EU law. The author presents a critical analysis, investigates the interaction between legal systems and suggests clarifications in order to attain the optimal functioning of the morality exclusion.

This thesis analyzes the current state of the selected Kindergarten. Based on the analysis made a proposal to extend kindergarten, due to lack of capacity building. The thesis is designed to address deficiencies occurring in kindergarten. It created a proposal to extend the new kindergarten class together with the calculation of the costs incurred and revenues.

L’objectif de ces travaux de thèse est de proposer des approches bayésiennes “physiquement informées” pour l’estimation de directions d’arrivée de sources acoustiques sous-marines dans un milieu océanique fluctuant. Dans un premier temps nous verrons donc comment décrire ces fluctuations comme perturbation locale de la célérité du milieu pour ensuite décrire leur impact sur la propagation d’une onde et la mesure de celle-ci. Ce faisant nous pourrons décrire la dégradation subie par le signal, motivant la création d’un modèle statistique pour décrire cette perturbation. Après un état de l’art sur les méthodes d’estimation des directions d’arrivées (DOA), tout d’abord en milieu constant, puis en milieu incertain, nous définirons une modélisation de ces fluctuations sous forme d’un bruit de phase structuré selon une distribution gaussienne multivariée. Nous verrons que ce choix nous rapproche du modèle théorique de structure défini dans le premier chapitre. Cette modélisation nous permet alors de proposer la méthode paSAMP, algorithme d’Approximate Message Passing, inspirée des méthodes de reconstruction de phase et se révélant plus robuste au bruit additif, ainsi qu’au bruit de phase structuré, que les méthodes d’estimation des DOA classiques. Dans un deuxième temps, nous proposerons de réviser notre modélisation statistique et de considérer un bruit de phase distribuée selon une loi de Von Mises multivariée. Cette distribution permet une modélisation plus fidèle des bruits de phase. Considérant ce nouveau modèle, nous dériverons deux nouvelles méthodes : VitAMin et VistaBEM, respectivement extensions de paSAMP et paVBEM, déjà présent dans la littérature et reposant sur une autre approximation variationnelle bayésienne. Les premiers résultats, réalisés sur des données synthétiques considérant un modèle plus simple de VonMises unidimensionnel, s’avèrent prometteurs quant à la bonne intégration des modèles considérés et ainsi à leur robustesse au bruit de phase distribué selon ce modèle. Pour finir nous verrons que par le traitement de données réelles, il nous est possible d’extraire les mêmes grandeurs caractéristiques que celles décrites par le modèle théorique, mais également d’identifier certaines ondes internes spécifiques. Ceci à des fins de caractérisation rapide des fluctuations du milieu par méthode passive ou encore pour une initialisation informée des algorithmes proposés
The purpose of these works is to propose new "physically informed" bayesian algorithms for DOA estimationin presence of random fluctuations due to the presence of internal waves in the propagation medium. First we will seehow to describe these fluctuations as a local perturbation of the celerity of the medium and then the impact of such waveson a propagated signal. In doing so, we are able to observe the corruption of this signal and to propose a statistical modelof this perturbation.After a state of the art of DOA estimation techniques in both constant an uncertain media, we will propose a modelizationof the fluctuations as a multiplicative phase noise following a multivariate gaussian distribution. Very close to thetheoretical model, this distribution allows us to develop the paSAMP algorithm, an Approximate Message Passingtechnique inspired form phase retrieval litterature which integrates this multiplicative phase noise model. In order to getcloser to the theoretical model, we propose to upgrade the phase noise prior using a multivariate Von Mises prior,allowing us to propose new bayesian methods for DOA estimation such as the VitAMin and the VistaBEM algorithms asextensions of resp. the paSAMP and the paVBEM algorithms. The first results on synthetic measurement considering aunivariate Von Mises phase noise seems are promising regarding the robustness of these algorithms to this new noisemodel.To conclude, we will see with the processing of real data from the ALMA campaign that it is possible to retrieve quantitiesfrom the theoretical model through signal statistics. Moreover, these quantities allow us to identify known phenomenon inthe propagation medium with only passive measurement. Such knowledge can then be use for tomography or as "smartinitialization" of the different algorithms

The solutions proposed in this thesis contribute to improve gait recognition performance in practical scenarios that further enable the adoption of gait recognition into real world security and forensic applications that require identifying humans at a distance. Pioneering work has been conducted on frontal gait recognition using depth images to allow gait to be integrated with biometric walkthrough portals. The effects of gait challenging conditions including clothing, carrying goods, and viewpoint have been explored. Enhanced approaches are proposed on segmentation, feature extraction, feature optimisation and classification elements, and state-of-the-art recognition performance has been achieved. A frontal depth gait database has been developed and made available to the research community for further investigation. Solutions are explored in 2D and 3D domains using multiple images sources, and both domain-specific and independent modality gait features are proposed.

In this study, an appearance reconstruction method based on extraction of material re&
#64258
ectance properties of a three-dimensional (3D) object from its twodimensional (2D) images is explained. One of the main advantages of this system is that the reconstructed object can be rendered in real-time with photorealistic quality in varying illumination conditions. Bidirectional Re&
#64258
ectance Distribution Functions (BRDFs) are used in representing the re&
#64258
ectance of the object. The re&
#64258
ectance of the object is decomposed into di&
#64256
use and specular components and each component is estimated seperately. While estimating the di&
#64256
use components, illumination-invariant images of the object are computed from the input images, and a global texture of the object is extracted from these images by using surface particles. The specular re&
#64258
ectance data are collected from the residual images obtained by taking di&
#64256
erence between the input images and corresponding illumination-invariant images, and a Lafortune BRDF model is &
#64257
tted to these data. At the rendering phase, the di&
#64256
use and specular components are blended into each other to achieve a photorealistic appearance of the reconstructed object.

La cryo-microscopie électronique est une technique tomographique permettant de reconstituer la structure 3D d’un objet complexe en biologie à partir d’un jeu d’acquisitions. Ces images de l’objet complexe sont appelées les projections et sont acquises sous orientations inconnues. Un des avantages de la cryo-microscopie électronique est l’obtention d’un modèle 3D de très haute résolution de l’objet dans un état naturel. La procédure de reconstruction comporte plusieurs étapes telles que l’alignement, la classification des projections, l’estimation de leurs orientations et le raffinement des projections. Lors de ces étapes, la distance entre deux projections est fréquemment mesurée. Le travail réalisé au cours de cette thèse s’organise autour de la recherche théorique d’une distance entre des projections non-orientées avec comme objectif l’amélioration de la procédure de reconstruction tomographique en cryo-microscopie électronique. La contribution de ce travail de thèse est une méthode permettant d’estimer la différence angulaire entre deux projections dans les cas 2D et 3D. Notre méthode est basée sur la construction d’un graphe de voisinage dont les sommets sont les projections, dont les arêtes relient des projections voisines et sont pondérées par une approximation locale de la différence angulaire. Le calcul de ces poids repose sur les propriétés des moments de projection. Notre méthode est testée sur des images simulées de différentes résolutions et de différents niveaux du bruit. La comparaison avec des autres méthodes d’estimation de la différence angulaire est aussi réalisée
Cryo-electron microscopy is a tomographic technique allowing to reconstruct a 3D model of complex structure in biology from a set of acquired images. These images are known as the tomographic projections and are taken at unknown directions. The advantage of the cryo-electron microscopy is the 3D reconstruction at very high resolution. The reconstruction procedure consists of many steps such as projection alignment, projection classification, orientation estimation and projection refinement. During these steps, the distance between two projections is frequently measured. The work in this thesis aims at studying the distances mesured between two unknown-direction projections with the objective of improving the reconstruction result in the cryo-electron microscopy. The contribution of this thesis is the developement of a method for estimating the angular difference between two projections in 2D and 3D. Our method is based on the construction of a neighborhood graph whose vertices are the projections, whose edges link the projection neighbors and are weighted by a local approximation of the angular difference. The calculation of the weights relies on the projection moment properties. The proposed method has been tested on simulated images with different resolutions and at different noise levels. The comparison with others estimation methods of angular difference has been realised

Esta tesis se centra en las reconstrucciones históricas en el cine europeo y su principal objetivo es delimitar la genealogía de un modelo de figuración y transmisión de la historia, enraizado en la modernidad y que emerge tras la Segunda Guerra Mundial, persistiendo en el cine y televisión contemporáneos. Dicho modelo tiene por eje central imágenes cuya puesta en escena arrastra al espectador al tiempo representado, que se figura en la pantalla como una experiencia sensorial y dialéctica gracias a la cual la audiencia puede efectuar una relectura de la historia. La aproximación háptica y dialéctica al pasado permite construir imágenes que operan como espacios de experiencia para el espectador, generándose una temporalidad que evoca la vivencia del tiempo histórico pero también genera una nueva visión del presente. A través de esta mirada a la historia y sus reconstrucciones cinematográficas exploramos las temporalidades que se generan tanto en la escritura histórica a través de la imagen como en su encuentro con la audiencia.
This thesis focuses on historical reconstructions in European cinema and its main objective is to define the genealogy of a model of figuration and transmission of history, rooted in modernity and emerging after World War II, persisting in contemporary film and television. This model has main axis images whose mise-en-scene drags the viewer to the time represented, thanks to figuring it as a sensory and dialectical experience that allows the audience to reread the past. The haptic and dialectical approach to the past allows to build images that operate as spaces of experience for the viewer, generating a temporality that evokes the experience of historical time, but also generates a new vision of the present. Through this look at history and its cinematic reconstructions we explore the temporalities that are generated, both in historical writing through images and in its encounter with the audience.

The Pierre Auger Observatory (PAO) was constructed to study the highest energy cosmic rays (UHECR). A hybrid of ground array and fluorescence detector, it is the largest ultra-high energy cosmic ray detector to date. As such, the PAO detects UHECR in unprecedented amounts, offering unique insights into the nature and origin of these most extraordinarily energetic particles in the universe. In this thesis we improve the accuracy of arrival direction uncertainty estimates for reconstructed events. We validate these improved uncertainty estimators through a number of statistical techniques, involving both recorded and simulated data. Furthermore, we identify and correct a number of systematic errors which arise in algorithmic corner cases. We propose novel techniques for measuring the time synchronisation of each detector in the using recorded air shower data. We use these techniques to measure the synchronisation of each detector across the PAO surface detector and fluorescence detector. Finally, we perform a cursory search for a point source of UHECR at the Galactic centre. A slight over-density of events is measured from the direction of the Galactic centre, however this over-density is not substantial enough to indicate a departure from isotropy.
Thesis (Ph.D.)--University of Adelaide, School of Chemistry and Physics, 2013

碩士
國立中正大學
電機工程研究所
104
Because of advancing with satellite technology in recent years, a lot of GPS techniques are also such rapid development and popularity that every car set up the GPS. However, most of the contents are based on 2D street maps or 3D graphic only in some importance scenes like viaducts and so on. These scenes are produced based on manual mode, not only much time-consuming but the quality of watching also is less realness. In order to impress users more realness, the GPS on image based has come to be the most trends which we take the pictures at the real scenes using them in the reconstruction techniques on image based. But GPS on image based needs a lot of manpower now, it build the models first, and then map the texture of buildings on the models. We propose a method which direction of shooting are the same as the driving direction with stereo camera. Using it reconstruct models of the 3D building cuboid and the most importance part is getting the orientation, length width, and height. First, we calibrate the stereo camera include intrinsic matrix, extrinsic matrix and fisheye distortion parameter. And then, after using Hough Transform to get the vanishing point of the buildings with parallel lines in the image, we can estimate the orientation of the cuboid models. We use the features on the skyline to get the lengths of the 3D cuboid models in our system. Finally, we use the particle filter to refine the 3d coordinates. Finally, we provide users surfing and navigation the 3D Buildings through Virtual Reality techniques to impress users.

博士
國立交通大學
機械工程系所
101
A particle-based quiet direct simulation (QDS) method [Albright et al., 2002] was invented to solve the inviscid Euler equation, in which the Maxwell-Boltzmann velocity distribution is enforced through the use of Gauss-Hermite quadrature integration without using any random number. It is a very fast Euler equation solver, which is deterministic with large dynamic range of flow properties like most conventional CFD methods, employs true-direction conservative fluxes for faithfully mimicking real flow motion, is highly localized (a small stencil) for easier parallelization and requires very low memory because the discrete velocities can be re-used in each cell. However, it is numerically very diffusive and has been extended to a nearly second-order numerical scheme by Smith et al. [2009] without really considering true-direction flux reconstruction. Thus, we intend to further address this problem from both numerical and theoretical viewpoints in this thesis. In the numerical part, a true-direction flux reconstruction of the second-order quiet direct simulation (QDS) as an equivalent Euler equation solver, called QDS-N2, is presented. Because of the true-directional nature of QDS, where volume-to-volume (true direction) fluxes are computed, as opposed to fluxes at cell interfaces as employed by traditional finite volume schemes, a volumetric reconstruction is required to reach a totally true-direction scheme. The conserved quantities are permitted to vary (according to a polynomial expression) across all simulated dimensions. Prior to the flux computation, QDS particles are introduced using properties based on weighted moments taken over the polynomial reconstruction of the conserved variables such as mass, momentum and energy. The resulting flux expressions are shown to exactly reproduce the existing second-order extension for a one-dimensional flow, while providing a means for true multi-dimension reconstruction.The new reconstruction is demonstrated in several verification studies. These include several two-dimensional test cases such as shock bubble interaction problem, an Euler-four-shock interaction, Euler-four-contact interaction, Mach 3 facing over a forward step, and the advection of a vortical disturbance. These results are presented, and the increased computational time and the effect of higher-order extension are discussed. The results show that the proposed multi-dimensional reconstruction provides a significant increase in the accuracy of the solution as compared to the previously developed QDS-2N method. We show that, despite the increase in the computational expense, the computational speed of the proposed QDS-N2 method is several times higher than that of the previously proposed QDS-2N method for a fixed degree of numerical accuracy, at least, for the test problem of the advection of vertical disturbances. As mentioned earlier, QDS method is intrinsically a highly localized numerical scheme, which makes it highly suitable for parallel computing on distributed-memory cluster machines using domain decomposition. With parallel implementation, an extension to three-dimensional QDS method is also demonstrated. The results show that the parallel efficiency, based on a strong scaling study, for a large-scale problem using 0.5, 2, and 12.5 million cells can reach up to 75%, 68.5%, and 65.5% with 256 processors respectively. In addition, the parallel efficiency, based on a weak scaling study, for a shock bubble interaction is 1.2, which the ideal efficiency is 1.0, up to 49 processors for 20,000 cells per processor. Note all the parallel performance tests were performed at the APLS cluster of National Center for High-Performance Computing, Taiwan. In the theoretical part, we have derived the analytical expressions of all the fluxes related to mass, momentum and energy in the two-dimensional QDS-N2 and QDS-2N methods respectively. Comparisons are made systematically between the corresponding fluxes in the two methods by varying flow properties. Results show that a large discrepancy of fluxes between these two methods occurs in the ranges of low density, low temperature, and high velocity.It is also interesting to learn that this range of gas flow often corresponds to an expansion wave region. Moreover, the fluxes using both methods are similar horizontal and vertical directions (x and y-direction), while large discrepancy is found in the fluxes going to the diagonal direction.With this observation, we can evaluate the accuracy of QDS-2Nmethod as compared to QDS-N2 method in the flow field, which may be important in deciding which method to be used for different flow problems. The major findings of the research with the recommendations for future study are summarized at the end of the thesis.

碩士
國立臺灣大學
資訊工程學研究所
90
A new method for human face reconstruction is developed in this work, which deforms a 3D generic face model based on multiple images taken from different viewing directions. The developed method consists of two stages. In the first stage, 3D positions of some features of the initial face model are reconstructed via a camera-pose-estimation technique under orthogonal projection. By applying an interpolation technique, the face model is deformed from the initial face model to fit the computed 3D feature points. In the second stage, an energy function is formulated to measure differences between the image contours and the projected contours of the 3D face model. The optimization process is performed in frequency domain for achieving global and smooth adjustment. With the developed technique, realistic face models can be generated.

碩士
國立臺北科技大學
土木與防災研究所
99
As the prompt development of IT technology, plus the continuous enhancement of camera resolution, utilizing photogrammetry to establish highly precise 3-dimensional (3D) models becomes more and more popular. Photogrammetry can also provide visible multi-angle measurement and display. Compare with traditional measurement methodology, photogrammetry is a great time saver. However, for different fields of photogrammetry applications such as bridge monitoring, slope monitoring, and plastic surgery … etc, the precision requirements of 3D models are not quite the same. The 3D-models precision requirements for the plastic surgery and surgical operation should at least up to millimeter level. Domestic literature indicates that object-space matching related studies have already successfully established facial 3D models with the precision up to less than 1mm. However, these studies lack of applying ear 3D models to the facial models which will cause failure to establish overall facial landmark. This is the enhance focus for this study. There are 5 parts in this study: (1) change from the past only taking straight facial photos with multi-viewing angles approach; (2) add in the concept of camera lens distortion difference to object-space matching, establish overall facial 3D model with left-side, middle, and right-side view; (3) by adding in two methodologies (product point cloud model set triangular mesh & apply orthophoto and adding in point cloud model to the RGB data) to display facial models; (4) utilizing 2D photogrammetry and 3D model measurement methodology to conduct 10 facial symbol landmarks measurements, and performance the comparison of the precision; (5) in the end of the experiment, will apply this facial symbol landmark 3D measurement to 30 grown ups (aged between 19-30) to establish statistics. The measuring results like overall facial line length and angles can be provided for further analysis and reference in the future.

碩士
國立暨南國際大學
資訊工程學系
90
Reconstruction of the 3-D shape of an object with high fidelity using stereo vision is always an important research topic in the field of computer vision. Human can effortlessly match the stereo correspondences of a point but are not skilful to accurately determine the 3-D coordinates of the point. On the other hand, a machine vision system can easily determine the 3-D coordinates of a point given correct stereo correspondences and the calibration parameters, but there is no efficient algorithm for determining stereo correspondences robustly. Because an interactive 3-D reconstruction system possesses both the merits of the human vision and the machine vision, stereo correspondences can be robustly estimated under the guidance of the operator yielding accurate 3-D reconstruction results. Therefore, lots of research groups have focused on developing interactive 3-D reconstruction system. However, the current human-computer interface is designed for word processing and 2-D graphics, thus is not suitable for interactive 3-D reconstruction. In this thesis, a bi-directional visual stereo interface for accessing the stereo matching results from a human brain is proposed. The stereo interface consists of a binocular gaze tracking system and a stereo display. The binocular gaze tracker comprises a binocular pupil tracker and a head tracker. A robust algorithm is developed to extract the pupil centers of user’s eyes in video rate (30 Hz). The estimated pupil positions plus the estimated head position can be used to provide 3-D lines of sight of user’s both eyes. While the operator stares at a point shown on the stereo display, the binocular gaze tracker records the stereo correspondences for guiding the interactive system to reconstruct 3-D information contained in the stereo images. Experimental results show that the proposed interface is very promising.

The purpose of the thesis was project a field road to reconstruct and modernize the current traffic connections between selected villages Sviny and Borkovice, using technology for application of Glorit. It was created reinforced major field road with turnouts the design category of P 4,5/30. Studies of the field communication were prepared in two different variants. The beginning of routes is in the village Sviny and end is in the village Borkovice. Both variants are connected to the current road. After the consultation with the executive of the thesis, one of the variants was evolved for building licence documentation. The documentation consists of plans and manual. Emphasis was placed mainly on routes to achieve fluency with regard to appropriate and sensitive integration of the field road to the countryside. There was also an effort to minimize earthwork and associated financial intensity of construction. The project of the field route was prepared in conformity with applicable national standards (DIN) and technical regulations that relate to the issue.