A sustainable society is a society that satisfies its resource requirements without endangering the sustainability of these resources. The mineral endowment on the African continent is estimated to be the first or second largest of world reserves. Therefore, it is recognised that the African continent still heavily depends on mineral exports as a key contributor to the gross domestic product (GDP) of various countries. These mining activities, however, do introduce primary and secondary environmental degradation factors. They attract communities to these mining areas, light and heavy industrial establishments occur, giving rise to artisanal activities. This study focussed on satellite RS products as an aid to a mine’s operations and the monitoring of its environment. Effective operational mine management and control ensures a more sustainable and profitable lifecycle for mines. Satellite based RS holds the potential to observe the mine and its surrounding areas at high temporal intervals, different spectral wavelengths and spatial resolutions. The combination of SAR and optical information creates a spatial platform to observe and measure the mine’s operations and the behaviour of specific land cover and land use classes over time and contributes to a better understanding of the mining activities and their influence on the environment within a specific geographical area. This study will introduce an integrated methodology to collect, process and analyse spatial information over a specific targeted mine. This methodology utilises a medium resolution land cover base map, derived from Landsat 8, to understand the predominant land cover types of the surrounding area. Using very high resolution mono- and stereoscopic satellite imagery provides a finer scale analysis and identifies changes in features at a smaller scale. Combining these technologies with the synthetic aperture radar (SAR) applications for precise measurement of surface subsidence or upliftment becomes a spatial toolbox for mine management. This study examines a combination of satellite remote sensing products guided by a systematic workflow methodology to integrate spatial results as an aid for mining operations and environmental monitoring. Some of the results that can be highlighted is the successful land cover classification using the Landsat 8 satellite. The land cover that dominated the Kolomela mine area was the “SHRUBLAND/GRASS” class with a 94% coverage and “MINE” class of 2.6%. Sishen mine had a similar dominated land cover characteristic with a “SHRUBLAND/GRASS” class of 90% and “MINE” class of 4.8%. The Pléiades time-series classification analysis was done using three scenes each acquired at a different time interval. The Sishen and Kolomela mine showed especially changes from the bare soil class to the asphalt or mine class. The Pléiades stereoscopic analysis provided volumetric change detection over small, medium, large and recessed areas. Both the Sishen and Kolomela mines demonstrated height profile changes in each selected category. The last category of results focused on the SAR technology to measure within millimetre accuracy the subsidence and upliftment behaviour of surface areas over time. The Royal Bafokeng Platinum tailings pond area was measured using 74 TerraSAR-X scenes. The tailings wall area was confirmed as stable with natural subsidence that occurred in its surrounding area due to seasonal changes of the soil during rainy and dry periods. The Chuquicamata mine as a large open pit copper mine area was analysed using 52 TerraSAR-X scenes. The analysis demonstrated significant vertical surface movement over some of the dumping sites. It is the wish of the researcher that this dissertation and future research scholars will continue to contribute in this scientific field. These contributions can only assist the mining sector to continuously improve its mining operations as well as its monitoring of the primary as well as the secondary environmental impacts to ensure improved sustainability for the next generation. / Environmental Sciences / M. Sc. (Environmental Science)

The problem facing the Tlokwe Local Municipality is that it is Constitutional and legislatively obliged to avoid and/or mitigate the impact of potential disasters within its boundaries, through the effective management of potential disaster risks and disasters. The lack of effective risk management tools is especially concerning in the context of the water catchment management of the Mooi River, which is the main water supply of the Tlokwe local Municipality. The Mooi River is exposed to many potential catchment related hazards that could affect the municipality of which the origins are both anthropogenic and natural. Although, many of the impacts on the catchment arise beyond the boundaries and the control of the municipality, this does not relieve the municipality from the responsibility to develop tools to manage the risks. The aim of the study is to assist in addressing the above stated problem through establishing the requirements for a conceptual model for an effective spatial information system that will assist the municipality in effectively managing the potential disaster risks and disasters that may arise in the Mooi River Catchment area that could impact on the Tlokwe Local Municiaplity. The aim gives rise to three-research questions that are formulated as research objectives that are used to identify the conceptual model requirements. The first is to identify and conceptualise the constitutional and legislative obligations in respect of disaster risk management in general and specifically those governing the disaster disk management in the water catchment area for the Mooi River. The study of this objective not only highlight the constitutional and legal obligations that the local municipality is subject to, but provides legislatives remedies that the local municipality can utilise to assist with disaster risks reduction. The second is to identify and conceptualise the generic hazards that are related to water catchment areas (including the related groundwater compartments) and those specific in the Mooi River catchment area. In this section, potential anthropogenic and natural hazards are listed, a methodology for risk and vulnerability analyses is provided, and a concise study of quaternary catchment C23D is provided. The third is to identify and conceptualize the requirements for an effective conceptual model of GIS for Disaster Risk Management in the Tlokwe Local Municipality. An overview of a GIS is provided. The essential components of a generic information system namely, people, software, hardware, procedures and processes, data and telecommunications or networks are discussed. The information and system requirements cumulating from the analyses of the three research questions, serve as the drivers of the goal, outcomes and transformation process of the system as well as the requirements for the conceptual model. In this section: a comparison of the Systems Development Life Cycle (SDLC), Framework for Applied Systems Thinking (FAST), problem solving and the phases addressed in this study; the identification of the conceptual model requirements; and a concise systems conceptualisation of an effective GIS is provided. As the, mini-dissertation focuses on the needs for a conceptual model, the additional activities required before the system can be implemented are identified and formulated as recommendation that provide the opportunity for future research. / M Development and Management, North-West University, Potchefstroom Campus, 2014

The problem facing the Tlokwe Local Municipality is that it is Constitutional and legislatively obliged to avoid and/or mitigate the impact of potential disasters within its boundaries, through the effective management of potential disaster risks and disasters. The lack of effective risk management tools is especially concerning in the context of the water catchment management of the Mooi River, which is the main water supply of the Tlokwe local Municipality. The Mooi River is exposed to many potential catchment related hazards that could affect the municipality of which the origins are both anthropogenic and natural. Although, many of the impacts on the catchment arise beyond the boundaries and the control of the municipality, this does not relieve the municipality from the responsibility to develop tools to manage the risks. The aim of the study is to assist in addressing the above stated problem through establishing the requirements for a conceptual model for an effective spatial information system that will assist the municipality in effectively managing the potential disaster risks and disasters that may arise in the Mooi River Catchment area that could impact on the Tlokwe Local Municiaplity. The aim gives rise to three-research questions that are formulated as research objectives that are used to identify the conceptual model requirements. The first is to identify and conceptualise the constitutional and legislative obligations in respect of disaster risk management in general and specifically those governing the disaster disk management in the water catchment area for the Mooi River. The study of this objective not only highlight the constitutional and legal obligations that the local municipality is subject to, but provides legislatives remedies that the local municipality can utilise to assist with disaster risks reduction. The second is to identify and conceptualise the generic hazards that are related to water catchment areas (including the related groundwater compartments) and those specific in the Mooi River catchment area. In this section, potential anthropogenic and natural hazards are listed, a methodology for risk and vulnerability analyses is provided, and a concise study of quaternary catchment C23D is provided. The third is to identify and conceptualize the requirements for an effective conceptual model of GIS for Disaster Risk Management in the Tlokwe Local Municipality. An overview of a GIS is provided. The essential components of a generic information system namely, people, software, hardware, procedures and processes, data and telecommunications or networks are discussed. The information and system requirements cumulating from the analyses of the three research questions, serve as the drivers of the goal, outcomes and transformation process of the system as well as the requirements for the conceptual model. In this section: a comparison of the Systems Development Life Cycle (SDLC), Framework for Applied Systems Thinking (FAST), problem solving and the phases addressed in this study; the identification of the conceptual model requirements; and a concise systems conceptualisation of an effective GIS is provided. As the, mini-dissertation focuses on the needs for a conceptual model, the additional activities required before the system can be implemented are identified and formulated as recommendation that provide the opportunity for future research. / M Development and Management, North-West University, Potchefstroom Campus, 2014

Thesis (MScAgric (Agricultural Economics))—University of Stellenbosch, 2005. / The study broadly assesses the usefulness of available digital spatial land resource data as a source for agritourism and ecotourism information by comparing the inventory of available spatial data sets for South Africa and the SADC region, with the needs for spatial data as derived from a literature study of travel motivations and demand determinants. Spatial land resource data have been collected, processed and stored for agricultural planning purposes, like land suitability assessment, agricultural production and infrastructural planning. Given a) the growth in agritourism and ecotourism, b) the more detailed information required by tourists to aid them during decision making processes like destination selection and c) the progress in information technology rendering access of information via the internet easier; the question arose whether the available land resource digital data can be processed to provide relevant tourism information on internet websites. Four tasks had to be done, namely: a) identifying tourists’ needs by means of a literature study on travel motivations and demand determinants; b) identifying the land resource data sets that could be processed into information to meet these identified needs; c) determining the accessibility of spatial information on internet tourism websites to potential agritourists and ecotourists, and d) exploring opportunities for adding value by looking at what information existing websites are offering in comparison with what can be obtained from repackaging the land resource data. Common ground was found between the spatial tourist information needs and the available spatial land resource data. This, coupled with the ability of combining meteorological and other humanmade environmental data in GIS modelling, suggests that repackaging land resource data seems to have the potential to offer useful tourism information in correspondence with confirmed tourist information needs.

Pour rendre compte du traitement de l’information visuo-spatiale, l’hypothèse dominante est celle du calepin visuo-spatial (Baddeley, 1986). Cependant l’approche multi-systèmes est actuellement remise en question et les théories fonctionnelles défendent l’idée d’un système mnésique unique. C’est dans cette perspective que se situe notre recherche qui s’organise autour de trois questions. La première consiste à définir les traitements visuo-spatiaux mis en œuvre lors de la manipulation d’informations visuo-spatiales. La seconde tente d’établir des profils comportementaux des traitements visuo-spatiaux caractéristiques du vieillissement normal et pathologique (Alzheimer). La troisième porte sur l’expression de la flexibilité cognitive chez des sujets âgés lors de la résolution des blocs de Corsi.Deux expériences ont été réalisées dans lesquelles les participants (sujets Alzheimer et contrôles) devaient résoudre la tâche de Corsi, dans sa version ordre direct (expérience 1) et dans ces versions ordre direct et indirect (expérience 2). Les données ont été exploitées en recourant à une méthodologie d’analyse de protocoles individuels. Dans la première expérience, nous nous sommes intéressée aux erreurs rencontrées. Dans la seconde étude, nous avons étendu nos analyses aux traitements visuo-spatiaux inhérents à ces erreurs. Les résultats obtenus ont permis de définir cinq traitements visuo-spatiaux (vectoriel, vectoriel partiel, identité stricte, approximatif et mixte), de mettre en évidence des profils comportementaux concernant le traitement vectoriel partiel et identité stricte pour les sujets âgés sains lors de la condition indirecte et de montrer l’expression de flexibilité cognitive spontanée lors des traitements visuo-spatiaux séquentiels. Nos analyses ont ainsi permis de proposer une approche différente du traitement de l’information visuo-spatiale en substituant le concept de traitement visuo-spatial à celui de calepin visuo-spatial. / Currently, to account visuospatial information processing, the dominant hypothesis is that of the visuospatial sketchpad (Baddeley, 1896). However, such models postulating the existence of several independent modules are being questioned and recent functional theories defend a single memory system. Our research therefore is organized in this perspective around three questions.The first one aim to define the different visuospatial processing when visuospatial information is needed. The second one is to attempt to establish behavioral profiles of the visuospatial processing by normal ageing people and pathological one (Alzheimer). The third question is to express the cognitive flexibility of elderly patients when solving the Corsi block-tapping task.Two experiments were conducted where participants (seniors affected by Alzheimer’s disease ‘AD’ and a control group of elderly) had to solve the Corsi block-tapping task, in the direct condition (experiment 1) and in the direct and indirect conditions (experiment 2). The data were analyzed using the methodology of individual protocols analysis. In the first experiment, we examined different errors encountered during the resolution of the task. In the second research we extended our analysis to visuospatial processing associated with these errors.We identified five major visuospatial processing: “the vector processing”, “partially vectorial”, “strict identity”, “approximate identity” and “mixed”, to highlight behavioral profiles on the ‘partially vectorial’ processing and ‘strict identity’ to the healthy elderly subjects in the indirect condition and show the expression of spontaneous flexibility in AD patients and normal subjects. Our analysis allowed us to propose a different approach that improve knowledge about both normal and pathological (here Alzheimer’s disease) ageing. Analyzing the subjects in action, conducts to speak in terms of visuospatial processing rather than visuospatial sketchpad.

La représentation d'image est au cœur de beaucoup d'algorithmes de vision par ordinateur. Elle intervient notamment dans des tâches de reconnaissance de catégories visuelles comme la classification ou la détection d'objets. Dans ce contexte, la représentation "sac de mot visuel" (Bag of Visual Words ou BoVW en anglais) est l'une des méthodes de référence. Dans cette thèse, nous nous appuyons sur ce modèle pour proposer des représentations d'images discriminantes. Dans la première partie, nous présentons une nouvelle approche simple et efficace pour prendre en compte des informations spatiales dans le modèle BoVW. Son principe est de considérer l'orientation et la longueur de segments formés par des paires de descripteurs similaires. Une notion de "softsimilarité" est introduite pour définir ces relations intra et inter mots visuels. Nous montrons expérimentalement que notre méthode ajoute une information discriminante importante au modèle BoVW et que cette information est complémentaire aux méthodes de l'état de l'art. Ensuite, nous nous focalisons sur la description de l'information couleur. Contrairement aux approches traditionnelles qui s'appuient sur des descriptions invariantes aux changements d'éclairage, nous proposons un descripteur basé sur le pouvoir discriminant. Nos expérimentations permettent de conclure que ce descripteur apprend automatiquement un certain degré d'invariance photométrique tout en surclassant les descripteurs basés sur cette invariance photométrique. De plus, combiné avec un descripteur de forme, le descripteur proposé donne des résultats excellents sur quatre jeux de données particulièrement difficiles. Enfin, nous nous intéressons à la représentation de la couleur à partir de la réflectance multispectrale des surfaces observées, information difficile à extraire sans instruments sophistiqués. Ainsi, nous proposons d'utiliser l'écran et la caméra d'un appareil portable pour capturer des images éclairées par les couleurs primaires de l'écran. Trois éclairages et trois réponses de caméra produisent neuf valeurs pour estimer la réflectance. Les résultats montrent que la précision de la reconstruction spectrale est meilleure que celle estimée avec un seul éclairage. Nous concluons que ce type d'acquisition est possible avec des appareils grand public tels que les tablettes, téléphones ou ordinateurs portables / Image representation is in the heart of many computer vision algorithms. Different computer vision tasks (e.g. classification, detection) require discriminative image representations to recognize visual categories. In a nutshell, the bag-of-visual-words image representation is the most successful approach for object and scene recognition. In this thesis, we mainly revolve around this model and search for discriminative image representations. In the first part, we present a novel approach to incorporate spatial information in the BoVW method. In this framework, we present a simple and efficient way to infuse spatial information by taking advantage of the orientation and length of the segments formed by pairs of similar descriptors. We introduce the notion of soft-similarity to compute intra and inter visual word spatial relationships. We show experimentally that, our method adds important discriminative information to the BoVW method and complementary to the state-of-the-art method. Next, we focus on color description in general. Differing from traditional approaches of invariant description to account for photometric changes, we propose discriminative color descriptor. We demonstrate that such a color description automatically learns a certain degree of photometric invariance. Experiments show that the proposed descriptor outperforms existing photometric invariants. Furthermore, we show that combined with shape descriptor, the proposed color descriptor obtain excellent results on four challenging data sets.Finally, we focus on the most accurate color representation i.e. multispectral reflectance which is an intrinsic property of a surface. Even with the modern era technological advancement, it is difficult to extract reflectance information without sophisticated instruments. To this end, we propose to use the display of the device as an illuminant while the camera captures images illuminated by the red, green and blue primaries of the display. Three illuminants and three response functions of the camera lead to nine response values which are used for reflectance estimation. Results show that the accuracy of the spectral reconstruction improves significantly over the spectral reconstruction based on a single illuminant. We conclude that, multispectral data acquisition is potentially possible with consumer hand-held devices such as tablets, mobiles, and laptops

Cette thèse s'inscrit dans un partenariat entre l'INRA et l'INRIA et dans le champs de l'extraction de connaissances à partir de bases de données spatiales. La problématique porte sur la caractérisation et la simulation de paysages agricoles. Plus précisément, nous nous concentrons sur des lignes qui structurent le paysage agricole, telles que les routes, les fossés d'irrigation et les haies. Notre objectif est de modéliser les haies en raison de leur rôle dans de nombreux processus écologiques et environnementaux. Nous étudions les moyens de caractériser les structures de haies sur deux paysages agricoles contrastés, l'un situé dans le sud-Est de la France (majoritairement composé de vergers) et le second en Bretagne (Ouest de la France, de type bocage). Nous déterminons également si, et dans quelles circonstances, la répartition spatiale des haies est structurée par la position des éléments linéaires plus pérennes du paysage tels que les routes et les fossés et l'échelle de ces structures. La démarche d'extraction de connaissances à partir de base de données (ECBD) mise en place comporte différentes étapes de prétraitement et de fouille de données, alliant des méthodes mathématiques et informatiques. La première partie du travail de thèse se concentre sur la création d'un indice spatial statistique, fondé sur une notion géométrique de voisinage et permettant la caractérisation des structures de haies. Celui-Ci a permis de décrire les structures de haies dans le paysage et les résultats montrent qu'elles dépendent des éléments plus pérennes à courte distance et que le voisinage des haies est uniforme au-Delà de 150 mètres. En outre différentes structures de voisinage ont été mises en évidence selon les principales orientations de haies dans le sud-Est de la France, mais pas en Bretagne. La seconde partie du travail de thèse a exploré l'intérêt du couplage de méthodes de linéarisation avec des méthodes de Markov. Les méthodes de linéarisation ont été introduites avec l'utilisation d'une variante des courbes de Hilbert : les chemins de Hilbert adaptatifs. Les données spatiales linéaires ainsi construites ont ensuite été traitées avec les méthodes de Markov. Ces dernières ont l'avantage de pouvoir servir à la fois pour l'apprentissage sur les données réelles et pour la génération de données, dans le cadre, par exemple, de la simulation d'un paysage. Les résultats montrent que ces méthodes couplées permettant un apprentissage et une génération automatique qui capte des caractéristiques des différents paysages. Les premières simulations sont encourageantes malgré le besoin d'un post-Traitement. Finalement, ce travail de thèse a permis la création d'une méthode d'exploration de données spatiales basée sur différents outils et prenant en charge toutes les étapes de l'ECBD classique, depuis la sélection des données jusqu'à la visualisation des résultats. De plus, la construction de cette méthode est telle qu'elle peut servir à son tour à la génération de données, volet nécessaire pour la simulation de paysage / This thesis is part of a partnership between INRA and INRIA in the field of knowledge extraction from spatial databases. The study focuses on the characterization and simulation of agricultural landscapes. More specifically, we focus on linears that structure the agricultural landscape, such as roads, irrigation ditches and hedgerows. Our goal is to model the spatial distribution of hedgerows because of their role in many ecological and environmental processes. We more specifically study how to characterize the spatial structure of hedgerows in two contrasting agricultural landscapes, one located in south-Eastern France (mainly composed of orchards) and the second in Brittany (western France, \emph{bocage}-Type). We determine if the spatial distribution of hedgerows is structured by the position of the more perennial linear landscape features, such as roads and ditches, or not. In such a case, we also detect the circumstances under which this spatial distribution is structured and the scale of these structures. The implementation of the process of Knowledge Discovery in Databases (KDD) is comprised of different preprocessing steps and data mining algorithms which combine mathematical and computational methods. The first part of the thesis focuses on the creation of a statistical spatial index, based on a geometric neighborhood concept and allowing the characterization of structures of hedgerows. Spatial index allows to describe the structures of hedgerows in the landscape. The results show that hedgerows depend on more permanent linear elements at short distances, and that their neighborhood is uniform beyond 150 meters. In addition different neighborhood structures have been identified depending on the orientation of hedgerows in the South-East of France but not in Brittany. The second part of the thesis explores the potential of coupling linearization methods with Markov methods. The linearization methods are based on the use of alternative Hilbert curves: Hilbert adaptive paths. The linearized spatial data thus constructed were then treated with Markov methods. These methods have the advantage of being able to serve both for the machine learning and for the generation of new data, for example in the context of the simulation of a landscape. The results show that the combination of these methods for learning and automatic generation of hedgerows captures some characteristics of the different study landscapes. The first simulations are encouraging despite the need for post-Processing. Finally, this work has enabled the creation of a spatial data mining method based on different tools that support all stages of a classic KDD, from the selection of data to the visualization of results. Furthermore, this method was constructed in such a way that it can also be used for data generation, a component necessary for the simulation of landscapes

Throughout the last decade, vehicle localization has been attracting significant attention in a wide range of applications, including Navigation Systems, Road Tolling, Smart Parking, and Collision Avoidance. To deliver on their requirements, these applications need specific localization accuracy. However, current localization techniques lack the required accuracy, especially for mission critical applications. Although various approaches for improving localization accuracy have been reported in the literature, there is still a need for more efficient and more effective measures that can ascribe some level of accuracy to the localization process. These measures will enable localization systems to manage the localization process and resources so as to achieve the highest accuracy possible, and to mitigate the impact of inadequate accuracy on the target application. In this thesis, a framework for fusing different localization techniques is introduced in order to estimate the location of a vehicle along with location integrity assessment that captures the impact of the measurement conditions on the localization quality. Knowledge about estimate integrity allows the system to plan the use of its localization resources so as to match the target accuracy of the application. The framework introduced provides the tools that would allow for modeling the impact of the operation conditions on estimate accuracy and integrity, as such it enables more robust system performance in three steps. First, localization system parameters are utilized to contrive a feature space that constitutes probable accuracy classes. Due to the strong overlap among accuracy classes in the feature space, a hierarchical classification strategy is developed to address the class ambiguity problem via the class unfolding approach (HCCU). HCCU strategy is proven to be superior with respect to other hierarchical configuration. Furthermore, a Context Based Accuracy Classification (CBAC) algorithm is introduced to enhance the performance of the classification process. In this algorithm, knowledge about the surrounding environment is utilized to optimize classification performance as a function of the observation conditions. Second, a task-driven integrity (TDI) model is developed to enable the applications modules to be aware of the trust level of the localization output. Typically, this trust level functions in the measurement conditions; therefore, the TDI model monitors specific parameter(s) in the localization technique and, accordingly, infers the impact of the change in the environmental conditions on the quality of the localization process. A generalized TDI solution is also introduced to handle the cases where sufficient information about the sensing parameters is unavailable. Finally, the produce of the employed localization techniques (i.e., location estimates, accuracy, and integrity level assessment) needs to be fused. Nevertheless, these techniques are hybrid and their pieces of information are conflicting in many situations. Therefore, a novel evidence structure model called Spatial Evidence Structure Model (SESM) is developed and used in constructing a frame of discernment comprising discretized spatial data. SESM-based fusion paradigms are capable of performing a fusion process using the information provided by the techniques employed. Both the location estimate accuracy and aggregated integrity resultant from the fusion process demonstrate superiority over the employing localization techniques. Furthermore, a context aware task-driven resource allocation mechanism is developed to manage the fusion process. The main objective of this mechanism is to optimize the usage of system resources and achieve a task-driven performance. Extensive experimental work is conducted on real-life and simulated data to validate models developed in this thesis. It is evident from the experimental results that task-driven integrity assessment and control is applicable and effective on hybrid localization systems.

Image representation is in the heart of many computer vision algorithms. Different computer vision tasks (e.g. classification, detection) require discriminative image representations to recognize visual categories. In a nutshell, the bag-of-visual-words image representation is the most successful approach for object and scene recognition. In this thesis, we mainly revolve around this model and search for discriminative image representations. In the first part, we present a novel approach to incorporate spatial information in the BoVW method. In this framework, we present a simple and efficient way to infuse spatial information by taking advantage of the orientation and length of the segments formed by pairs of similar descriptors. We introduce the notion of soft-similarity to compute intra and inter visual word spatial relationships. We show experimentally that, our method adds important discriminative information to the BoVW method and complementary to the state-of-the-art method. Next, we focus on color description in general. Differing from traditional approaches of invariant description to account for photometric changes, we propose discriminative color descriptor. We demonstrate that such a color description automatically learns a certain degree of photometric invariance. Experiments show that the proposed descriptor outperforms existing photometric invariants. Furthermore, we show that combined with shape descriptor, the proposed color descriptor obtain excellent results on four challenging data sets.Finally, we focus on the most accurate color representation i.e. multispectral reflectance which is an intrinsic property of a surface. Even with the modern era technological advancement, it is difficult to extract reflectance information without sophisticated instruments. To this end, we propose to use the display of the device as an illuminant while the camera captures images illuminated by the red, green and blue primaries of the display. Three illuminants and three response functions of the camera lead to nine response values which are used for reflectance estimation. Results show that the accuracy of the spectral reconstruction improves significantly over the spectral reconstruction based on a single illuminant. We conclude that, multispectral data acquisition is potentially possible with consumer hand-held devices such as tablets, mobiles, and laptops

Throughout the last decade, vehicle localization has been attracting significant attention in a wide range of applications, including Navigation Systems, Road Tolling, Smart Parking, and Collision Avoidance. To deliver on their requirements, these applications need specific localization accuracy. However, current localization techniques lack the required accuracy, especially for mission critical applications. Although various approaches for improving localization accuracy have been reported in the literature, there is still a need for more efficient and more effective measures that can ascribe some level of accuracy to the localization process. These measures will enable localization systems to manage the localization process and resources so as to achieve the highest accuracy possible, and to mitigate the impact of inadequate accuracy on the target application. In this thesis, a framework for fusing different localization techniques is introduced in order to estimate the location of a vehicle along with location integrity assessment that captures the impact of the measurement conditions on the localization quality. Knowledge about estimate integrity allows the system to plan the use of its localization resources so as to match the target accuracy of the application. The framework introduced provides the tools that would allow for modeling the impact of the operation conditions on estimate accuracy and integrity, as such it enables more robust system performance in three steps. First, localization system parameters are utilized to contrive a feature space that constitutes probable accuracy classes. Due to the strong overlap among accuracy classes in the feature space, a hierarchical classification strategy is developed to address the class ambiguity problem via the class unfolding approach (HCCU). HCCU strategy is proven to be superior with respect to other hierarchical configuration. Furthermore, a Context Based Accuracy Classification (CBAC) algorithm is introduced to enhance the performance of the classification process. In this algorithm, knowledge about the surrounding environment is utilized to optimize classification performance as a function of the observation conditions. Second, a task-driven integrity (TDI) model is developed to enable the applications modules to be aware of the trust level of the localization output. Typically, this trust level functions in the measurement conditions; therefore, the TDI model monitors specific parameter(s) in the localization technique and, accordingly, infers the impact of the change in the environmental conditions on the quality of the localization process. A generalized TDI solution is also introduced to handle the cases where sufficient information about the sensing parameters is unavailable. Finally, the produce of the employed localization techniques (i.e., location estimates, accuracy, and integrity level assessment) needs to be fused. Nevertheless, these techniques are hybrid and their pieces of information are conflicting in many situations. Therefore, a novel evidence structure model called Spatial Evidence Structure Model (SESM) is developed and used in constructing a frame of discernment comprising discretized spatial data. SESM-based fusion paradigms are capable of performing a fusion process using the information provided by the techniques employed. Both the location estimate accuracy and aggregated integrity resultant from the fusion process demonstrate superiority over the employing localization techniques. Furthermore, a context aware task-driven resource allocation mechanism is developed to manage the fusion process. The main objective of this mechanism is to optimize the usage of system resources and achieve a task-driven performance. Extensive experimental work is conducted on real-life and simulated data to validate models developed in this thesis. It is evident from the experimental results that task-driven integrity assessment and control is applicable and effective on hybrid localization systems.