This thesis examines the evolution of the now dominant contour method of relief representation, associated interpretation problems and attempted solutions. A hybrid method of relief representation is then suggested combining a standard topographic map, to provide quantitative data, and computer generated three-dimensional visualisations, to give topographic structure. The hypothesis that this system will result in improved performance at relief interpretation tasks, particularly for inexperienced map users, was then tested. A second aspect of the hypothesis centred on the colouring of the 3-D views. Thus the experiment was expanded to assess if the more natural impression provided by layer colouring resulted in improved performance. The principle evaluation involved 2 Objective Tests of user performance. Subjective assessment of 3-D visualisation production and quality and Opinion Samples were also used in the overall evaluation. This thesis concludes that the three-dimensional information tested here has not produced the expected improvement. Reasons suggested for this include lack of training with 2 and 3 dimensional data, unfamiliarity with computer generated three-dimensional views and inability of inexperienced map users to manipulate complex three-dimensional imagery. Suggested improvements to the theory include increased training and use of more specifically targeted views, for example, to illustrate a walking route using a series of views along its course.

Les ursidés (Mammalia, Carnivora, Ursidae) représentent une part importante du registre fossile au Pléistocène moyen et supérieur. Pour l’ours des cavernes (Ursus deningeri– Ursus spelaeus), de nombreux débats s’articulent autour de son régime alimentaire. Des caractères dentaires, tels que les dimensions et le relief occlusal, suggèrent une alimentation végétarienne exclusive, mais de récents travaux argumentent en faveur d’un spectre alimentaire plus omnivore. Les dents d’ours des cavernes et brun (Ursus arctos) montrent aussi une importante variabilité morphologique difficile à quantifier et à interpréter par la morphométrie traditionnelle. Une analyse des contours via la transformée de Fourier elliptique a été appliquée sur les dents jugales d’ursidés fossiles et actuels afin d’étudier la variabilité de forme, en complément d’une analyse traditionnelle. Un test préliminaire a d’abord été effectué sur une population d’U. spelaeus (grottes de Montaigle, Belgique). Puis, les séquences de Sous-Scladina et Scladina (Sclayn, Belgique) ont été analysées afin d’observer les variabilités de taille et de forme au cours du temps. Certains ursidés actuels, tels le panda géant Ailuropoda melanoleuca et l’ours blanc Ursus maritimus, sont spécialisés, tandis que d’autres sont plus généralistes (i. e. l’ours brun). Après avoir appliqué la méthode sur les ursidés actuels, une approche actualiste a été testée en intégrant des ursidés fossiles provenant de plusieurs localités. Certains caractères indiqueraient une alimentation plus végétarienne pour l’ours brun archaïque et plus variée pour l’ours des cavernes. / Ursid (Mammalia, Carnivora, Ursidae) remains are often found in abundance in the fossil records of Middle and Upper Pleistocene. For the extinct cave bear (Ursus deningeri – Ursus spelaeus), many debates currently focus on its diet. Dental features such as the dimensions and the occlusal relief suggest an exclusively vegetarian diet, but recent studies indicate a more omnivorous range food. Cave and brown bear (Ursus arctos) teeth also show a strong morphological variability, difficult to quantify and to interpret using traditional morphometrics. To complete these traditional analyses, an outline analysis via the elliptic Fourier transform has been performed on fossil and extant bear cheek teeth to assess the shape variability. Firstly, a population of U. spelaeus (Montaigle Caves, Belgium) has been tested. Then, the sequences of Sous-Scladina and Scladina Caves (Sclayn, Belgium) have been considered to assess size and shape modifications through time. Some extant bears, such as the giant panda Ailuropoda melanoleuca or the polar bear Ursus maritimus show a high diet specialization, whereas some extant ursids are generalist (i. e. the brown bear). After applying the method on the main extant taxa to establish an actual framework, an actualist approach has been finally tested by adding fossil bears from several localities. Few dental features suggest a more vegetarian archaic brown bear, whereas the cave bear diet would be more diversified than presumed.

This study discusses the three-dimensional refraction of progressive wave trains propagating over a bottom of circular concentric contours and the results are expressed in a polar coordinate. First, a general differential formulation of refraction is derived via three different methods: by transferring from its original Cartesian form to the polar coordinate, by applying the Fermat¡¦s principle in polar coordinate, and by applying the conservation of waves in polar coordinate. All three approaches give the same governing equation; hence, its correctness is verified. Based on this governing equation, the wave ray, the phase function, the constant phase line, and the refraction coefficient are all determined. In the present refraction problem for an originally uniform wave train propagating over a bottom of circular concentric contours, a few special features, including the cusps of constant phase lines due to the effect of bottom, and the envelope composed of these cusps, are present. All these refraction properties can be expressed in terms of both a snapshot and a time evolution of constant phase lines. In the lee side of the shoal, there exists a sheltered zone that is enclosed by the envelope of the cusps. In this zone, wave rays intersect and the corresponding caustic problem arises, and all possible combinations of intersecting rays are also specifically described in this study. The difficulty of classical ray theory for the caustic problem is overcome and the caustic phenomenon and its refraction coefficients are determined explicitly in this study.

This thesis develops image segmentation methods for the application of automated cervical cancer screening. The traditional approach to automating this task has been to emulate the human method of screening, where every one of the hundreds of thousands of cells on each slide is analysed for abnormality. However, due to the complexity of cervical smear images and the low error tolerance imposed upon the segmentation stage, only limited success has previously been found. A different approach is to detect malignancy associated changes (MACs) in a relatively small sample of the total population of cells. Under this paradigm, the requirement to segment every cell is loosened, but delineation accuracy and error checking become essential. Following a review of generic and cervical smear segmentation, it is concluded that prior work on the traditional approach to automation is not suitable for a MACs solution. However, the previously proposed framework of a dual-magnification system is found to be relevant and is therefore adopted. Here, scene images are first captured at low resolution in order to rapidly locate the cells on a slide. Cells that are deemed to be suitable for further analysis are then imaged at high resolution for the more accurate segmentation of their nuclei. A water immersion algorithm is developed for low resolution scene segmentation. This method achieves a rapid and robust initial segmentation of the scene without the requirement of incorporating extensive a priori knowledge of the image objects. A global minimum searching contour is presented as a top-down method for segmenting the high resolution cell nucleus images where the image objects are well characterised by shape and appearance. This latter method is tested upon 20,000 images and found to achieve an accurate segmentation rate of 99.47%. An error checking method, that uses segmentation stability as an indicator of segmentation success, is developed that is capable of detecting 100% of the failures of the nucleus segmenter, at the expense of discarding only 9% of the data. Throughout this work, contemporary issues in the field of generic image segmentation are presented and some of these are addressed for the cervical smear application. Finally, an avenue of future work is proposed which may lead to the much wider proliferation of computer vision solutions to everyday problems.

The purpose of the present study was to tease apart the paralinguistic from the linguistic aspects of infants' perception of infant-directed (ID) speech. Several studies have shown that infants beginning at a few days after birth discriminate native from nonnative speech and can discriminate specific contours (rising, falling, rising-falling) in ID speech. Some studies have also indicated that infants at 4.5 months of age prefer their own name over other names but at 6 months of age, fail to prefer a sentence with their own name embedded in it. Using a discrimination procedure, the current study investigated whether 6-month-old infants could detect a change in contour and/or a change in words when listening to ID utterances. Results indicated that 6-month-old infants detected both a contour and a word change. From these results, it is argued that as has been shown in other experiments, infants are extremely sensitive to subtle changes in speech. Furthermore, ID speech appears to facilitate infants' ability to discriminate small changes in ID speech (both linguistic and paralinguistic). It is suggested that future studies investigate more discrete changes in speech samples and a replication of this research with adult-directed (AD) speech. / Master of Science

No / The tilt aftereffect (TAE) has been used previously to probe whether contours defined by different attributes are subserved by the same or by different underlying mechanisms. Here, we compare two types of contours between texture surfaces, one with texture orientation contrast across the edge (orientation contrast contour; OC) and one without, commonly referred to as a subjective contour (SC). Both contour types produced curves of TAE versus adapting angle displaying typical positive and negative peaks at ~15 and 70 deg, respectively. The curves are well fit by difference of Gaussian (DoG) functions, with one Gaussian accounting for the contour adaptation effect and the other accounting for the texture orientation adaptation effect. Adaptation to OC elicited larger TAEs than did adaptation to SC, suggesting that they more effectively activate orientation-selective neurons in V1/V2 during prolonged viewing. Surprisingly, both contour types adapted a luminance contour (LC) as strongly as did an LC itself, suggesting that the second-order orientation cue contained in the texture edge activates the same set of orientation-selective neurons as does an LC. These findings have implications for the mechanisms by which the orientations of texture edges and SCs are encoded

Cette thèse concerne la segmentation d’images HRpQCT et l’évaluation d’indices morphologiques de l’os cortical pour le diagnostic de l’ostéoporose et la prédiction des fractures osseuses. Dans un premier temps,deux méthodes sont proposées pour la segmentation de l’os cortical. La première utilise une nouvelle approche des contours actifs basée sur la logique floue suivie d’une nouvelle technique de remplissage développée pour imiter le comportement des opérateurs pour séparer l’os cortical de l’os trabéculaire. La deuxième approche est une technique 3D à double contours actifs combinant à la fois les informations locales le long et entre les deux contours. Les deux approches de segmentation sont comparées à celles de l’état de l’art afin de valider leurs performances. Dans un second temps, différents indices extraits de l’os cortical sont utilisés pour déterminer leur potentiel de prédiction des fractures ostéoporotiques. Les résultats obtenus montent que l’analyse globale de l’os cortical masque des variations potentiellement importantes.Par conséquent, une décomposition régionale de l’enveloppe corticale est proposée afin d’améliorer la prédiction du risque fracturaire. / This thesis concerns the segmentation of HRpQCT images and the evaluation of the cortical bone parameters for the osteoporosis characterization and the fracture prediction. Firstly, two approaches were proposed to segment the cortical bone. The first uses a new fuzzy energy active contours approach followed by a new filling technique designed to mimic the behaviour of clinicians while extracting the cortical bone from the trabecularone. The second approach is a local based 3D dual active contours approach proposed to separate between three regions constituting the image. To move, this approach combines the local information along each point in the two contours conjointly with the information between them. The segmentation results of these approaches were confronted to the state of the art methods to validate their performance. Secondly,different parameters were extracted from the segmented cortical bone to monitor the association of these parameters with the osteoporotic fracture prediction. Global analysis of the cortical bone obscures potentially important regional variations. Therefore, regional cortical decomposition was proposed to illustrate that cortical sub-regions could improve the evaluation of fracture risk than the global analysis of the cortical bone.

Les fissures et les nids-de-poule sont des défauts très présents sur les routes du réseau routier québécois. Un bon contrôle de la qualité lors de la pose de l’enrobé bitumineux permet de diminuer les risques d’apparition de ces défauts. Le ministère des Transports du Québec (MTQ) utilise la thermographie infrarouge afin de détecter les zones non conformes, soit celles qui deviendront des nids-de-poule ou des fissures. Des variations thermiques sur l’image infrarouge permettent la détection de ces zones. Toutefois, les logiciels utilisés par le MTQ ne sont pas appropriés pour détecter les zones non conformes. Ce mémoire présente deux méthodes de détection automatique des zones non conformes. La première permet l’analyse des images prises par une caméra thermique alors que la seconde permet d’analyser en continu les données provenant d’un scanneur infrarouge. Ces deux méthodes utilisent des techniques de segmentation afin de détecter les zones non conformes. Elles permettent l’analyse automatique des données sans qu’aucune intervention humaine ne soit nécessaire.

Maps of runoff-depth have been found to be useful tools in a variety of water resource applications. Producing such maps can be a challenging and expensive task. One of the standard methods of producing these maps is to use a manual procedure based on gaged runoff data, topographic and past runoff-depth maps, and the expert opinion of hydrologists. This thesis examined five new automated procedures for producing runoff-depth contour maps to see if the maps produced by these procedures had similar accuracy and characteristics when compared to the manual procedure. An uncertainty analysis was used to determine the accuracy of the automated procedure maps by withholding gaged runoff data from the creation of the contour maps and then interpolating estimated runoff back to these sites from the maps produced. Subtracting gaged runoff from estimated runoff produced interpolation error values. The mean interpolation error was used to define the accuracy of each map and was then compared to a similar study by Rochelle, et al., (1989) conducted on a manual procedure map.

Here we search for the curve which has the smallest integral of the square of curvature, while passing through two given points with given orientation. This is the true shape of a spline used in lofting. In computer-aided design, curves have been sought which maximize "smoothness". The curve discussed here is the one arising in this way from a commonly used measure of smoothness. The human visual system may use such a curve when it constructs a subjective contour.