The objective of the present work is to propose a novel method to extract a stable feature set representative of image content. Each image is represented by a linear combination of fractal orthonormal basis vectors. The mapping coefficients of an image projected onto each orthonormal basis constitute the feature vector. The set of orthonormal basis vectors are generated by utilizing fractal iterative function through target and domain blocks mapping. The distance measure remains consistent, i.e., isometric embedded, between any image pairs before and after the projection onto orthonormal axes. Not only similar images generate points close to each other in the feature space, but also dissimilar ones produce feature points far apart. The above statements are logically equivalent to that distant feature points are guaranteed to map to images with dissimilar contents, while close feature points correspond to similar images. Therefore, utilizing coefficients derived from the proposed linear combination of fractal orthonormal basis as key to search image database will retrieve similar images, while at the same time exclude dissimilar ones. The coefficients associated with each image can be later used to reconstruct the original. The content-based query is performed in the compressed domain. This approach is efficient for content-based query. Scaling, rotational, translation, mirroring and horizontal/vertical flipping variations of a query image are also supported. A symbolic image database system is a system in which a large amount of image data and their related information are represented by both symbolic images and physical images. How to perceive spatial relationships among the components in a symbolic image is an important criterion to find a match between the symbolic image of the scene object and the one being store as a modal in the symbolic image database. Spatial reasoning techniques have been applied to pictorial database, in particular those using 2D strings as an index representation have been successful. In most of the previous approaches for iconic indexing, for simplifying the concerns, they apply the MBR (Minimum bounding rectangle) of two objects to define the spatial relationship between them. Multiple instance learning algorithms provide ways for computer program to improve automatically with experience. Most images are inherently ambiguous disseminators of information. Unfortunately, interfaces to image databases normally involve the user giving the system ambiguous queries. By treating each query as a Multiple-Instance example, we make the ambiguity in each image explicit. In addition, by receiving several positive and negative examples, the system can learn what the user desires. Using the learned concept, the system returns images from the database that are close to that concept. In this project, we propose to apply the Multiple-Instance learning model by deriving the projection vector of fractal orthonormal bases for a small number of training images to learn what images from database are of interest to the user.

This thesis tackles the problem of automatically synthesizing motions of close-character interactions which appear in animations of wrestling and dancing. Designing such motions is a daunting task even for experienced animators as the close contacts between the characters can easily result in collisions or penetrations of the body segments. The main problem lies in the conventional representation of the character states that is based on the joint angles or the joint positions. As the relationships between the body segments are not encoded in such a representation, the path-planning for valid motions to switch from one posture to another requires intense random sampling and collision detection in the state-space. In order to tackle this problem, we consider to represent the status of the characters using the spatial relationship of the characters. Describing the scene using the spatial relationships can ease users and animators to analyze the scene and synthesize close interactions of characters. We first propose a method to encode the relationship of the body segments by using the Gauss Linking Integral (GLI), which is a value that specifies how much the body segments are winded around each other. We present how it can be applied for content-based retrieval of motion data of close interactions, and also for synthesis of close character interactions. Next, we propose a representation called Interaction Mesh, which is a volumetric mesh composed of points located at the joint position of the characters and vertices of the environment. This raw representation is more general compared to the tangle-based representation as it can describe interactions that do not involve any tangling nor contacts. We describe how it can be applied for motion editing and retargeting of close character interaction while avoiding penetration and pass-throughs of the body segments. The application of our research is not limited to computer animation but also to robotics, where making robots conduct complex tasks such as tangling, wrapping, holding and knotting are essential to let them assist humans for the daily life.

Of fundamental concern in volcanic hazard and risk assessment studies of volcanic systems is what role crustal structures might play in the ascent of magma through the crust. What are the processes that govern the spatial distribution and timing of eruptions, especially in populated areas or near sensitive facilities? Many studies have drawn the conclusion that faults play a critical role as easily–exploitable crustal weaknesses along which magma can ascend. Great care must be used when assuming a causative relationship between patterns of vents and faults especially when such relationships may be incorporated into hazard assessment models or other forecasting tools. This thesis presents a quantitative analysis of vent and fault populations in seven actively–faulted volcanic fields to test whether or not spatial relationships exist between faults and volcanic features. The data generated in this study include map distances acquired by measuring existing geologic maps produced by other scientists. Statistical methods were adapted from a similar study by Paterson and Schmidt (1999) which involved the analysis of pluton–to–fault distances. The data show that statistical spatial correlations exist between vents and faults in only two of the seven volcanic fields in this study. As a general observation, most vents cluster far from faults in these populations, which could be explained by a variety of natural phenomenon such as suppression of faulting from increased magmatism and magma source geometry differences. Although data some of the data show a spatial correlation, it does not necessarily imply a genetic relationship.

In this thesis, we propose a new representation, which we name Interaction Bisector Surface (IBS), that can describe the general nature of spatial relationship. We show that the IBS can be applied in 3D scene analysis, retrieval and synthesis. Despite the fact that the spatial relationship between different objects plays a significant role in describing the context, few works have focused on elaborating a representation that can describe arbitrary interactions between different objects. Previous methods simply concatenate the individual state vectors to produce a joint space, or only use simple representations such as relative vectors or contacts to describe the context. Such representations do not contain detailed information of spatial relationships. They cannot describe complex interactions such as hooking and enclosure. The IBS is a data structure with rich information about the interaction. It provides the topological, geometric and correspondence features that can be used to classify and recognize interactions. The topological features are at the most abstract level and it can be used to recognize spatial relationships such as enclosure, hooking and surrounding. The geometric features encode the fine details of interactions. The correspondence feature describes which parts of the scene elements contribute to the interaction and is especially useful for recognizing character-object interactions. We show examples of successful classification and retrieval of different types of data including indoor static scenes and dynamic scenes which contain character-object interactions. We also conduct an exhaustive comparison which shows that our method outperforms existing approaches. We also propose a novel approach to automatically synthesizing new interactions from example scenes and new objects. Given an example scene composed of two objects, the open space between the objects is abstracted by the IBS. Then, an translation, rotation and scale equivariant feature called shape coverage feature, which encodes how the point in the open space is surrounded by the environment, is computed near the IBS and around the open space of the new objects. Finally, a novel scene is synthesized by conducting a partial matching of the open space around the new objects with the IBS. Using our approach, new scenes can be automatically synthesized from example scenes and new objects without relying on label information, which is especially useful when the data of scenes and objects come from multiple sources.

It is well-known that, when it comes to crime, some neighborhoods are safer than others. Researchers who make maps of crime have observed that some areas of cities have more crime than others. These areas of high crime are often called hot spots. Crime pattern theory explained why some neighborhoods have more crime than others by looking at criminal events as a meeting between a motivated criminal and a target. Social scientists, geographers, and city planners have shown that criminals generally choose targets from places they see every day, for example on their ride to work or the grocery store. This means that when the daily routine of a criminal changes, the location of that person's criminal behavior could change too. When trends in the daily routine of a whole city change, the location of crimes in that city could change because criminals, in general, will choose targets from different places in the city. In fact, some researchers have suggested that crimes will become clustered around transportation nodes, such as street car stations, after new lines are opened. But so far only a few studies have tried to demonstrate the pattern hot spots follow in the years following major transportation changes. The answer to this question is important to urban designers and police because it would allow them to respond to changes in the location of hot spots when new public transportation projects occur.

Within the context of a site/place on which a skyscraper is developed, the role of the corporate plaza is usually limited to providing a base to glorify the unique form of the skyscraper (the office tower). In such case, the potential for a symbiotic relationship between the skyscraper and the plaza is not realized. In exploring the possibilities of a symbiotic relationship the first step was to research the skyscraper and its evolution. This analysis based on existing literature revealed the 'motive' behind the design and construction of skyscrapers and the reason for its strong physical image - "the corporations desire for an attention drawing device". This desire has resulted in the design of the skyscraper as an object in the urban landscape with a high degree of 'recognizability'. Although the skyscraper's character of recognizability plays an important role in the perception of the site, it fails to provide for the development of 'experiential meaning'. This experiential meaning which is a critical component for the environmental image of the site/place can be provided for in the plaza space. In order for the meaning, that is developed at the plaza, to be associated with the skyscraper, the skyscraper and the plaza should have a sense of continuity and appear as pats of a whole. To achieve this sense of continuity, the design of the plaza space needs to reflect/incorporate the elements that contribute to the strong physical image of the skyscraper as means of reinforcing the image of the skyscraper that is stored in the memory. The remainder of the thesis was to apply and test the above statement/idea in a design solution. the first step in that process was the analysis of existing plazas against the statement-case studies. the case studies reinforced a few aspects that needed to be addressed in the design of the plaza 1) reference to the imageable character can be made through the use of materials as much as shape and forms (and there might be other possibilities based on the character of the building) 2) The functional aspect of the plaza is critical for the successful use of the plaza, an assumption that supports this thesis. Once inferences were made from the case studies the next step was to decide on a site to redesign a plaza space- The PPG place in Pittsburgh. The PPG place, a complex of 6 buildings, exemplifies the postmodern movement and was designed by one of the chief proponents of the movement, Philip Johnson. The once stark and desolate plaza was refurbished recently with the addition of fountains around the obelisk in the center. A skating rink is operated in the plaza during the winter months. The design process was initiated by three layers of analysis 1) Analysis of the imageable qualities of the building 2) Analysis of the issues pertaining to the usability of the plaza 3) Analysis of the urban context These three layers of information provided a broad framework for the design of the plaza, from which the final design concept was distilled. the space that is thus designed facilitates a high degree of usability as well as a relationship to the imageability of the skyscraper, ensuring a holistic image for the site/place in which the skyscraper and plaza coexist in a mutualistic relationship. / Master of Landscape Architecture

Cette thèse a pour objectif de développer un système de reconnaissance de structures logique des documents d'entreprises sans modèle. Il s'agit de reconnaître la fonction logique de blocs de textes qui sont importants à localiser et à identifier. Ce problème est identique à celui de la détection d'objets dans une scène naturelle puisqu'il faut à la fois reconnaître les objets et les localiser dans une image. A la différence de la reconnaissance d'objets, les documents d'entreprises doivent être interprétés sans aucune information a priori sur leurs modèles de structures. La seule solution consiste à développer une approche non supervisée basée principalement sur les relations spatiales et sur les informations textuelles et images. Les documents d'entreprises possèdent des contenus et des formes très hétérogènes car chaque entreprise et chaque administration créent son propre formulaire ou ses propres modèles de factures. Nous faisons l'hypothèse que toute structure logique de document est constituée de morceaux de micro-structures déjà observées dans d'autres documents. Cette démarche est identique en détection d'objets dans les images naturelles. Tout modèle particulier d'objet dans une scène est composé de morceaux d'éléments déjà vu sur d'autres exemples d'objets de même classe et qui sont reliés entre eux par des relations spatiales déjà observées. Notre modèle est donc basé sur une reconnaissance partie par partie et sur l'accumulation d'évidences dans l'espace paramétrique et spatial. Notre solution a été testée sur des applications de détection d'objets dans les scènes naturelles et de reconnaissance de structure logique de documents d'entreprises. Les bonnes performances obtenues valident les hypothèses initiales. Ces travaux contiennent aussi de nouvelles méthodes de traitement et d'analyse d'image couleurs de documents et d'images naturelles. / This digital revolution introduces new services and new usages in numerous domains. The advent of the digitization of documents and the automatization of their processing constitutes a great cultural and economic revolution. In this context, computer vision provides numerous applications and impacts our daily lives and businesses. Behind computer-vision technology, fundamental concepts, methodologies, and algorithms have been developed worldwide in the last fifty years. Today, computer vision technologies arrive to maturity and become a reality in many domains. Computer-vision systems reach high performance thanks to the large amount of data and the increasing performance of the hardware. Despite the success of computer-vision applications, however, numerous other applications require more research, new methodologies, and novel algorithms. Among the difficult problems encountered in the computer-vision domain, detection remains a challenging task. Detection consists of localizing and recognizing an object in an image. This problem is far more difficult than the problem of recognition alone. Among the numerous applications based on detection, object detection in a natural scene is the most popular application in the computer-vision community. This work is about the detection tasks and its applications.

Thesis (MA)--Stellenbosch University, 2000. / ENGLISH ABSTRACT: European cities offer a variety of attractions, activities and facilities to tourists. One component of tourism in this urban context is cultural tourism which is a high potential growth sector within the tourism industry. Cultural tourists travel to cities mainly to see the most important and well known cultural attractions, to take part in cultural festivities, to view the cultural treasures of the cities and to attend performances and exhibitions. Other functions and activities such as retail functions and types also occur in the vicinity of the important cultural attractions of a city. Amongst these activities and functions are shops, cafés and restaurants. In an area of a city where cultural attractions and retail types coexist, the possibility exits that a spatial relationship between cultural tourism and retailing may occur in the sense that the tourists who visit these attractions may support the retail establishments in the vicinity of cultural attractions. The city of Leuven in Belgium is a typical West European city with a variety of cultural tourist attractions as well as functions which can be supportive of cultural tourism. These functions include shops, cafés and restaurants. An investigation was conducted to establish whether a spatial relationship exists between cultural tourism and retail activities in the part of Leuven known as the Ring. The geographical information systems TransCad, Arc/Info en ArcView were used for the processing and analysis of retail and cultural tourism data gathered in the Ring of Leuven. Three analytical procedures were performed separately for shops, cafés and restaurants. These procedures comprised (a) the determination of the numbers and types of supportive functions around the most important cultural-historical tourist attractions, (b) the determination of the numbers and types of supportive functions along the routes to cultural-historical sites from tourist points of origin, and (c) the determination of the numbers and types of supportive functions along guided-walk routes in the city. The analysis results show that a (spatial) relationship does exist between retail and cultural tourism in the Ring of Leuven. On average 44% of the retail functions (shops, cafés and restaurants) are located within a distance of up to saam from the six most important cultural building attractions in Leuven (procedure 1). Retail functions on guided-walk routes totalled 48% with cafés and restaurants being the most abundant (procedure 3). Only 19% of retail functions are located between points of origin and cultural building attractions (procedure 2). A more detailed study should be conducted to establish the exact nature of the spatial relationship which exists between retail and cultural tourism in the Ring of Leuven. / AFRIKAANSE OPSOMMING: Europese stede bied 'n reeks attraksies, aktiwiteite en fasiliteite vir toeriste. Een komponent van die breër term toerisme in hierdie stedelike konteks is kulturele toerisme wat 'n baie sterk groeisektor binne die toerismemark is. Kulturele toeriste besoek stede hoofsaaklik met die doelom die belangrikste en bekendste kulturele attraksies te besigtig, om aan kulturele feeste deel te neem, die kunsskatte van stede te betrag en om teateropvoerings en tentoonstellings by te woon. In die nabyheid van die belangrikste kulturele attraksies van 'n stad kom ook ander funksies en aktiwiteite soos kleinhandelsfunksies en -tipes voor. Hieronder ressorteer onder andere winkels, cafés en restaurante. In die gedeelte van 'n stad waar kulturele attraksies en kleinhandelstipes naby mekaar voorkom, kan die moontlikheid bestaan van 'n ruimtelike verwantskap tussen kulturele toerisme en kleinhandel, in dié sin dat toeriste wat die attraksies besoek, moontlik ook die kleinhandelstipes wat in die nabyheid van die kulturele attraksies voorkom, salondersteun. Die stad Leuven in België is 'n tipiese voorbeeld van 'n Wes-Europese stad met 'n verskeidenheid kulturele toeriste-attraksies asook funksies wat kulturele toerisme ondersteun. Hierdie ondersteunende funksies sluit winkels, cafés en restaurante in. In die deel van Leuven wat as die Ring bekend staan, is ondersoek ingestel na die voorkoms, al dan nie, van 'n ruimtelike verwantskap tussen kulturele toerisme en kleinhandel. Die geografiese inligtingstelsels TransCad, Arc/Info en ArcView is ingespan vir die verwerking en analise van data oor kleinhandel en kulturele toerisme in die Ring van Leuven. Drie ontledingsprosedures is afsonderlk uitgevoer vir winkels, cafés en restaurante. Hierdie prosedures behels (a) die bepaling van die aantal en tipes ondersteunende funksies rondom die belangrikste kultuurhistoriese toeristeattraksies; (b) die bepaling van die aantal en tipe ondersteunende funksies langs die roetes tussen die toeriste-oorsprongpunte en die belangrikste kultuurhistoriese toeriste-attraksies; en (c) die bepaling van die aantal en tipe ondersteunende funksies langs begeleide stadswandelroetes. Die analiseresultate wys dat 'n verwantskap tussen kleinhandel en kulturele toerisme, ruimtelik-gesproke, wel in die Ring van Leuven bestaan. Só is gemiddeld 44% van die kleinhandelsfunksies (winkels, cafés en restaurante) tot op In afstand van SOOmvanaf die ses belangrikste kulturele gebou-attraksies in Leuven geleë (prosedure 1). Ook kom 48% van die kleinhandelsfunksies langs die roetes van begeleide stadswandelinge voor. Hier is dit egter die cafés en restaurante wat die grootste proporsies uitmaak (prosedure 3). 'n Gemiddelde proporsie van slegs 19% van die kleinhandelsfunksies kom tussen oorsprongpunte en die belangrikste kulturele gebou-attraksies voor. 'n Meer omvattende studie moet onderneem word om vas te stel watter tipe ruimtelike verwantskap tussen kleinhandel en kulturele toerisme binne Leuven se Ring voorkom.

Les habitats côtiers à fonds meubles sont essentiels dans le cycle de vie de nombreuses espèces de poissons. Ces habitats ont pour caractéristique d’être peu étendus mais sont aussi très productifs et jouent à ce titre un rôle de nourricerie pour les juvéniles de poissons bentho-démersaux. Les fortes abondances de proies semblent être l’une des raisons principales de la présence des juvéniles au sein de ces habitats. Bien que déjà étudié, le caractère limitant de la nourriture disponible fait toujours l’objet de débats dans la communauté scientifique. Une des raisons principales est que l’étude des milieux côtiers est rendue complexe par la diversité des facteurs entrant en jeu dans le développement des jeunes stades de poissons. A partir du cas d’étude de la baie de Vilaine, une des nourriceries les plus productives du golfe de Gascogneles recherches menées dans cette thèse visent à définir le rôle du facteur alimentaire dans l’organisation de la nourricerie et dans sa capacité à soutenir le développement des juvéniles de poissons. Le fil conducteur de cette thèse est donc de déterminer si le facteur alimentaire limite la production de juvéniles. Les résultats montrent une forte exploitation de la production alimentaire ainsi qu'une superposition spatiale entre les densités de juvéniles de poissons et de leurs proies. La cohérence de ces résultats tend à soutenir l'hypothèse que la capacité d'accueil de la baie de Vilaine est atteinte et donc que le facteur alimentaire est limitant. Les implications de ce mécanisme de régulation sur la dynamique des populations nourricer / Soft sediment coastal habitats are essential in the life cycle of numerous fishes. These habitats are spatially-limited but very productive, and play a key role of nursery for the juveniles of benthic and demersal fishes. High abundance of prey seems to be one of the main reasons of the presence of juvenile fish within these habitats. Although widely studied, the limiting aspect of the feeding factor is still debated in the scientific community. One of the main reasons is that studying coastalhabitats is complex because numerous factors influence the development of the first life stages of fish. From the study case of the Bay of Vilaine, one of the most productive nurseries of the Bay of Biscay, research conducted in this thesis aims to define the role of the feeding factor in the organisation of the nursery and in its capacity to support the development of juvenile fish. The common thread of this thesis is thus to determine if the feeding factor limits the juvenile fish production. The results show a strong exploitation of the food production by the juvenile fish community and a spatial match between the densities of juvenile fish and their prey. The consistency of these results tends to support that the hypothesis that the carrying capacity of the Bay of Vilaine is reached, and that the feeding factor is limiting. The consequences of this regulation process on the dynamics of nursery-dependent fish populations are discussed.