<p> Sedimentary structures, textures and fabric were studied in detail in eight turbidite greywacke beds from the Cloridorme Formation (Middle Ordovician), Gaspe, Quebec. The beds are exposed on the wave-cut platform near Grande Vallee and were traced for about two miles along the strike (276 degrees), which is parallel to the average current direction (274 degrees) as indicated by sole mark directions. </p> <p> The beds under study show the following systematic downcurrent changes along the strike from east to west: (i) Sole mark types show the following changes: (1) tool marks such as grooves, prod marks and isolated flute marks, to (2) longitudinal ridges, longitudinal ridges with overlapping flute, closely spaced flutes, to (3) poorly developed, shallow longitudinal ridges with occasional cuspate crossing bars, to (4) smooth bottom with occasional grooves. (ii) There is an increase in variance of sole mark directions without a large change in the mean direction in most beds, but the sole mark direction changes by nearly 90° as one of the beds is traced from east to west along the strike. (iii) Most of the beds are massive and are divided into two parts by a bedding joint. However, in the proximal region, if a bed is unusually thin, the bedding joint is absent and the bed shows slightly wavy, plane-laminated structure at places. (iv) Grain orientations are generally statistically non-significant in the upper massive part of the bed in the proximal region and significant in the rest of the bed. Significant grain orientations and graptolite orientations show large deviations from the sole in the distal region. Deviations increase towards the top of the bed. (v) The beds show a change from good grading to poor grading or slight reverse grading from the proximal to distal region. It is considered that the beds were deposited by low concentration, highly turbulent currents. Deposition of the beds took place in two distinct phases. First a 'quick' bed separated from the current was sheared extensively by the overflowing current and on consolidation formed the lower part of the bed. Later deposition in the proximal region from the upper part of the current and the part close to the tail was by settling of individual particles from the suspension. Separation and consolidation of another 'quick' bed in some cases formed the upper part of the beds in the distal region. </p> <p> The depositing currents produced a good vertical and lateral grain size grading in the proximal region, but poor vertical grading in the distal region due to the loss of coarse grains and possibly increased concentration of flocculated clay. Also, the turbidity currents tended to "meander" greatly on slowing down in the distal region. </p> / Thesis / Doctor of Philosophy (PhD)

On présente dans ce qui suit une étude détaillée de la performance du LACIF (locally adaptive contrast invariant filter), qui est à la base un filtre non linéaire invariant sous transformation linéaire d'intensité de la forme± af(x, y) + bO(x, y) ¿, où± f(x, y) ¿est l'objet à reconnaitre et± O(x, y) ¿ son support. ± a ¿ et ± b ¿ sont respectivement les facteurs multiplicatif et additif du changement d'intensité. Ainsi l'étude de performance est établie sous différents aspects et par le biais de différents outils qu'on a adopté tout au long de ce mémoire.

The Enåsen gold deposit is located in Gävleborg county in central Sweden. Mining operations at Enåsen took place from 1984 to 1991 with Au as the main target for exploitation. The deposit has been interpreted as a metamorphosed Palaeoproterozoic analogue to near-recent epithermal Au mineralisations of a high sulphidation type. Its present mineralogy, textural-structural features, and morphology have been suggested to be the result of a combination of later deformation and regional Svecokarelian metamorphism at upper amphibolite to granulite facies conditions of the original epithermal deposit and it’s hydrothermally altered host rock. The main ore body now consists of a mineralised sillimanite quartzite gneiss. The aim of the project was to characterise the ore mineralogy, petrography and its paragenesis, evaluate the potential of Enåsen in terms of critical metals, and to test a hypothesis of partial ore melting.Among the most frequent ore minerals in the deposit are pyrite, chalcopyrite, pyrrhotite, bornite and tetrahedrite-tennantite, with variable but less abundant sulphides including covellite, digenite, mawsonite, stannite, arsenopyrite, cobaltite, galena, marcasite, sphalerite and pentlandite. Additionally, native gold, Se-bearing tellurobismuthite, hessite, tsumoite, pilsenite, rickardite, vulcanite, altaite, molybdenite, frohbergite, montbrayite, tellurantimony, löllingite and tellurbismuthantimony. While not an ore mineral here, rutile occurs abundantly. The ore mineral assemblages have seemingly at least partially melted. This is evidenced by failed quenching textures in the form of abundant multi-scale symplectites, potential sharp dihedral angles, localised concentrations of low melting point chalcophile elements (LMCE) + Au and Ag and arrays of multiphase sulphide/sulphosalt ± gold inclusions, as well as available mineral stability data considering that the ore assemblages have been subjected to upper amphibolite/granulite facies conditions followed by ductile and brittle deformation stages. Some ore mineral relationships have been described. Further studies would be required for a full paragenesis. The potential of Enåsen type deposits in terms of critical or near critical metals and semi-metals is likely to be as biproduct extraction in a mining operation aimed at gold. The most relevant element is likely to be bismuth, followed by tellurium and antimony.

Perceptual learning occurs because observers become more sensitive to informative aspects of the stimuli. Learning the informative aspects of one stimulus set does not transfer to another stimulus set of the same class. In this dissertation, the argument will be made that if observers learn how to discover informative aspects, learning will be more generalizable. However, discovery requires that the informative aspects are not easily apparent. To this end, stimulus orientation structure can be manipulated to contain informative structure in one orientation band, and non-informative structure in the other orientation band. Such a manipulation was inspired by research on face perception: Faces are best identified when decisions are based more on the horizontal relative to the vertical facial structure. Hence, the first three chapters focus on understanding the horizontal bias during face identification, and the final two chapters introduce a novel stimulus set for which horizontal bias may be learned. Chapter 2 identifies a neural marker of horizontal bias that is correlated with face identification accuracy, suggesting that we can predict how well observers identify faces based on their neural sensitivity to horizontal relative to vertical structure. Chapter 3 shows that when face identification accuracy declines due to healthy ageing, so too do behavioural and neural horizontal bias, but Chapter 4 shows that perceptual learning can increase horizontal bias in healthy older adults. Chapter 5 uses texture stimuli and shows that observers can learn to discover informative horizontal structure embedded in uninformative vertical structure. Chapter 6 extends these findings to show that adequate practice results in learning that generalizes to novel textures for which the orientation-selective processing is relevant. The results presented inform our understanding of the neural representations associated with orientation-selective processing, and suggest that observers can learn to discover informative structure conveyed by a particular orientation band. / Thesis / Doctor of Philosophy (PhD)

Titre de l'écran-titre (visionné le 25 mars 2024) / Ce mémoire présente une nouvelle approche automatisée pour détecter les fuites thermiques lors des inspections de bâtiments. Nous exposons un cadre intégrant de manière cohérente l'enregistrement automatique d'images, la segmentation d'images et la fusion des résultats segmentés issus à la fois d'images visibles et thermiques. Notre recherche compare rigoureusement les techniques d'enregistrement manuel et automatique d'images, démontrant que la méthode automatique, exploitant l'algorithme de la Transformée de Caractéristiques Invariantes à l'Échelle (SIFT), surpasse l'approche manuelle en termes d'efficacité, de précision et de reproductibilité. La segmentation des images utilise une approche supervisée pour les images visibles, en adoptant spécifiquement une architecture UNET++ et Resnet 152, tandis que les images thermiques sont segmentées en utilisant la méthode non supervisée de Kanezaki. Les résultats segmentés de ces divers types d'images sont ultérieurement fusionnés, fournissant une représentation exhaustive et précise des fuites thermiques à travers différentes parties du bâtiment. La qualité de la fusion dépend de la précision des étapes antérieures, soulignant l'importance de chaque étape dans ce système intégré. Ce travail démontre l'efficacité de notre nouvelle approche dans le domaine de l'inspection automatisée des bâtiments et ouvre la voie à de futures explorations et innovations dans ce domaine. Le mémoire est structuré en quatre chapitres, débutant par une revue exhaustive de la littérature, suivie de la méthodologie proposée, des résultats et des discussions, et se conclut par les orientations pour les travaux futurs.

The search for mineral deposits is a time consuming, risky and very expensive process. Applying new models and methods provides a competitive advantage in the search for mineral deposits because an explorationist can quickly evaluate potential targets and eliminate areas without good potential for mineralization. This dissertation presents a practical technique for prediction to finding precious metal mineralization at Guanajauato mining district (GMD) base on theoretical and experimental studies of fluids properties, mineral phase equilibrium, physical and chemical mechanisms. Making the technique highly transportable so that it can be applied in the field during an exploration program base on petrographic characteristics of mineral textures produced during boiling events in quartz, calcite and adularia, and fluid inclusions contained in these phases. While this work was conducted in GMD, the results should be applicable in exploration for epithermal deposits worldwide. The GMD is one of the largest silver producing districts in the world. Ore shoots are localized along three major northwest trending vein systems, the La Luz, Veta Madre and Vetas de la Sierra. More than 1200 samples were collected from surface outcrops, underground mine and historical and recent drill core. Traverses perpendicular to veins in all system were also conducted. Most of the samples (approximately 90%) were also assayed for Au, Ag, Cu, Pb, Zn, As, Sb. Samples from the GMD show a wide range in silica textures are indicative of rapid precipitation, such as occurs when fluids boil. Other mineral phases, including illite, rhombic adularia and bladed calcite are also indicative of rapid growth in a hydrothermal system and are characteristic of boiling systems. Because boiling is an effective mechanism for precipitating gold and silver from hydrothermal fluids, the presence of mineral textures indicative of boiling is a desirable feature in exploration. In many samples, textural evidence for boiling is supported by coexisting liquid-rich and vapor-rich fluid inclusions, or Fluid Inclusion Assemblages consisting of only vapor-rich inclusions, suggesting "flashing" of the hydrothermal fluids. Textural and fluid inclusion evidence for boiling has been observed in the deepest levels of the GMD, suggesting that additional precious metal resources may occur beneath these levels. / Ph. D.

Planting date, seeding rate, soil texture, and row configuration are important factors in soybean production. Each of these factors can impact overall production and yield of soybean immensely. Growers can have difficulty making decisions about how to best manage their production systems with these factors in mind. Therefore, research was conducted from 2019 to 2021 at the Delta Research and Extension Center in Stoneville, MS, to evaluate the agronomic performance, yield components, and weed control of soybean planted with different planting dates, row configurations, soil textures and/or seeding rates. Common row configurations utilized in Mississippi soybean production were compared to a triple-row configuration on raised beds. First planting dates occurred from late-April to early-May and second planting dates followed three weeks later. Each row configuration was planted at 320,000 seed ha-1 for agronomic studies. Seeding Rate Study was initiated where triple-row configuration plots were planted at 320,000, 445,000, and 553,000 seed ha-1. Total dry matter (TDM) was determined by removing a m-2 in each plot at soybean growth stage R6.5 and allowing samples to dry down and weigh them. Harvest index was determined by collecting seed from TDM samples and weighing them. Pod node-1, seed number, and node plant-1 were determined by collecting five random plants from each plot and counting the total number of pods, seed, and nodes. Soybean planted on silt loam at the first planting produced lower seed weight, seed number, and harvest index than soybean on clay soil at either planting date. Two row configurations, single- and triple-row, were planted on raised beds for weed control studies. Programs included PRE only, EPOST, LPOST, PRE fb EPOST, PRE fb LPOST, and PRE fb EPOST fb LPOST. Herbicide timings included 7, 14, 21, and 28 DA-crop emergence and each timing was followed by a sequential application 14 d after the initial application. Triple-row configuration had lower plant densities and produced lower yield than single- and twin-row configurations. Triple-row configuration soybean planted on clay soil and at 445,000 and 553,000 seed ha-1 produced greater yield than any other seeding rate or soil texture.

En s’emparant de la question ornementale, le cinéma s’inscrit dans une réflexion reliant les territoires balkaniques de l’art, au-delà de l’opposition entre les beaux-arts, les arts décoratifs et le design. La pensée de l’ornement au cinéma ne relève pas de l’évidence, tant il est vrai que l’histoire de la notion est marquée par sa relégation au seuil de la création artistique. Pourtant, il semble bien que l’ornement soit présent partout où il faut combler du vide, décorer un objet, embellir le corps. À quoi tient l’attrait de l’ornement au cinéma ? Quelles sont les relations qu’entretient le cinéma avec la pensée ornementale des images ? Contre le préjugé qui frappe l’ornement d’insignifiance, il semble nécessaire d’éclairer ses implications morales, sociales et politiques. À travers un corpus circonstancié qui se déploie à l’entour du cinéma, de l’histoire de l’art et de la culture visuelle, ce travail de recherche entend cartographier les enjeux critiques, formels et perceptifs du régime visuel de l’ornement. Penser l’ornement aujourd’hui, c’est opérer un basculement dans l’ordre des valeurs dont il se fait l’héritier. Il s’agit non seulement de faire retour sur son histoire et son élaboration conceptuelle, mais également d’emmener l’ornement vers un au-delà du discours de la subsidiarité pour établir une pensée dynamique de l’ornemental au cinéma. À travers différentes propositions théoriques, se formulent autant de constellations articulant les enjeux esthétiques, historiques et politiques de l’ornement : le régime de la dépense et le corps féminin en exercice, le processus de réification de « l’ornement de la masse » dans la modernité, le discours sur la couleur ornementale et la hantise cosmétique, les formes naturelles et abstraites, le paradigme du tapis et la texture des images, les états altérés de la perception. Les questions ornementales adressées au cinéma s’intègrent dans une pensée rénovée du style – point vif de la tension qui anime l’art et la vie – en tant que procédure de qualification des formes. / As it seized the ornamental issue, cinema has become part of a reflection process relating the scattered territories of arts, beyond the opposition between fine arts, decorative arts and design. Thinking the ornament in cinema is far from obvious, since this notion, throughout its evolution, has always been marginalized at the threshold of artistic creation. Yet, it does seem that the ornament is present wherever a void needs to be filled, an object to be adorned or a body to be embellished. What makes the ornament appealing in cinema? What relationships does the cinema hold with the ornamental visual thinking? To counter prejudices confining the ornament within insignificance, it appears necessary to clarify its moral, social and political implications. Through a detailed corpus spreading around cinema, history of art and visual culture, this research paper aims at mapping the critical, formal and perceptual issues of the visual regime of the ornament. Thinking the ornament today means swaying the order of the values it has inherited. It is not only about looking back on its history and its conceptual elaboration, but also about bringing the ornament beyond the subsidiary speech to establish a dynamic thinking of the ornamental in cinema. By way of different theoretical suggestions, many constellations are being elaborated. They connect the aesthetic, historical and political stakes of the ornament: the regime of consumption and the gestures of the feminine body, the process of reifying the “mass ornament” into modernity, the speech on decorative colour and the cosmetic obsessive fear, the natural and abstract forms, the carpet paradigm and the texture of images, the altered states of the perception of ornaments. The ornamental issues, which are addressed to the cinema, blend into a renovated thinking on style – central in the tension that stirs arts and life – as a process of qualifying forms.

<p>The progression of technical development has yielded practicable camera systems for the acquisition of so called depth maps, images with depth information. </p><p>Images and movies with depth information open the door for new types of applications in the area of computer graphics and vision. That implies that they will need to be processed in all increasing volumes.</p><p>Increased depth image processing puts forth the demand for a standardized data format for the exchange of image data with depth information, both still and animated. Software to convert acquired depth data to such videoformats is highly necessary. </p><p>This diploma thesis sheds light on many of the issues that come with this new task group. It spans from data acquisition over readily available software for the data encoding to possible future applications. </p><p>Further, a software architecture fulfilling all of the mentioned demands is presented. </p><p>The encoder is comprised of a collection of UNIX programs that generate MPEG Z/Alpha, an MPEG2 based video format. MPEG Z/Alpha contains beside MPEG2's standard data streams one extra data stream to store image depth information (and transparency). </p><p>The decoder suite, called TexMPEG, is a C library for the in-memory decompression of MPEG Z/Alpha. Much effort has been put into video decoder parallelization, and TexMPEG is now capable of decoding multiple video streams, not only in parallel internally, but also with inherent frame synchronization between parallely decoded MPEG videos.</p>

Dans cette thèse, nous nous intéressons à un récent algorithme de segmentation d’images basé sur un processus de régularisation de graphes. L’objectif d’un tel algorithme est de calculer une fonction indicatrice de la segmentation qui satisfait un critère de régularité ainsi qu’un critère d’attache aux données. La particularité de cette approche est de représenter les images à l’aide de graphes de similarité. Ceux-ci permettent d’établir des relations entre des pixels non-adjacents, et ainsi de procéder à un traitement non-local des images. Afin d’en améliorer la précision, nous combinons cet algorithme à une seconde approche non-locale : des caractéristiques de textures. Un nouveau terme d’attache aux données est dans un premier temps développé. Inspiré des travaux de Chan et Vese, celui-ci permet d’évaluer l’homogénéité d’un ensemble de caractéristiques de textures. Dans un second temps, nous déléguons le calcul de l’attache aux données à un classificateur supervisé. Entrainé à reconnaitre certaines classes de textures, ce classificateur permet d’identifier les caractéristiques les plus pertinentes, et ainsi de fournir une modélisation plus aboutie du problème. Cette seconde approche permet par ailleurs une segmentation multiclasse. Ces deux méthodes ont été appliquées à la segmentation d’images texturées 2D et 3D. / In this thesis, we improve a recent image segmentation algorithm based on a graph regularization process. The goal of this method is to compute an indicator function that satisfies a regularity and a fidelity criteria. Its particularity is to represent images with similarity graphs. This data structure allows relations to be established between similar pixels, leading to non-local processing of the data. In order to improve this approach, combine it with another non-local one: the texture features. Two solutions are developped, both based on Haralick features. In the first one, we propose a new fidelity term which is based on the work of Chan and Vese and is able to evaluate the homogeneity of texture features. In the second method, we propose to replace the fidelity criteria by the output of a supervised classifier. Trained to recognize several textures, the classifier is able to produce a better modelization of the problem by identifying the most relevant texture features. This method is also extended to multiclass segmentation problems. Both are applied to 2D and 3D textured images.