Geopolymer cements present a unique opportunity to make concrete binders almost entirely out of waste stream materials. Geopolymers made from fly ash, a waste product of coal power generation, as the aluminosilicate source and caustic activating solution were the focus of this study. However, the use of waste stream materials presents many challenges. One major stumbling block is that fly ash is inherently variable in composition and difficult to comprehensively characterize. The purpose of this work was to clarify the relationship between fly ash composition and reactivity in geopolymer cements. Ten fly ashes comprising a wide compositional spectrum were selected for the study and were characterized using quantitative x-ray diffraction and multispectral image analysis (MSIA) of x-ray maps coupled with point compositional analysis. The fly ashes were mixed into geopolymer mortars to determine their reactivity when activated as geopolymers. I hypothesized that the fly ashes that performed well under geopolymer formation conditions would have similarities in the glassy phases identified in them. The fly ashes that resulted in geopolymers with high compressive strengths did have several glassy phases in common. The phases were typically high in calcium, high in silicon, and somewhat low in aluminum. To determine whether the common phases were soluble and therefore likely to be dissolved, a dissolution method was used in which fly ash was mixed with concentrated caustic solution and continuously agitated; after 7 d and 28 d, the solid residues from the dissolution were studied using MSIA. The results showed that most of the glassy phases hypothesized to react were reactive, although the results were somewhat complex due to the heterogeneity of fly ash. The MSIA method proposed in previous work was further developed through this study, and a new way of selecting the training classes for phase composition assignment in the images was proposed. / text

Imagens multiespectrais estão presentes na grande maioria de dispositivos de imageamento atuais, desde câmeras pessoais até microscópios, telescópios e satélites. No entanto, grande parte dos trabalhos em análise de texturas e afins propõem abordagens monocromáticas, que muitas vezes consideram apenas níveis de cinza. Nesse contexto e considerando o aumento da capacidade dos computadores atuais, o uso da informação espectral deve ser considerada na construção de modelos melhores. Ultimamente redes neurais convolucionais profundas pré-treinadas tem sido usadas em imagens coloridas de 3 canais, porém são limitadas a apenas esse formato e computam muitas convoluções, o que demanda por hardware específico (GPU). Esses fatos motivaram esse trabalho, que propõem técnicas para a modelagem e caracterização de imagens multiespectrais baseadas em redes complexas, que tem se mostrado uma ferramenta eficiente em trabalhos anteriores e possui complexidade computacional similar à métodos tradicionais. São introduzidas duas abordagens para aplicação em imagens coloridas de três canais, denominadas Rede Multicamada (RM) e Rede Multicamada Direcionada (RMD). Esses métodos modelam todos os canais da imagem de forma conjunta, onde as redes possuem conexões intra e entre canais, de forma parecida ao processamento oponente de cor do sistema visual humano. Experimentos em cinco bases de textura colorida mostram a proposta RMD supera vários métodos da literatura no geral, incluindo redes convolucionais e métodos tradicionais integrativos. Além disso, as propostas demonstraram alta robustez a diferentes espaços de cor (RGB, LAB, HSV e I1I2I3), enquanto que outros métodos oscilam de base para base. Também é proposto um método para caracterizar imagens multiespectrais de muitos canais, denominado Rede Direcionada de Similaridade Angular (RDSA). Nessa proposta, cada pixel multiespectral é considerado como um vetor de dimensão equivalente à quantidade de canais da imagem e o peso das arestas representa sua similaridade do cosseno, apontando para o pixel de maior valor absoluto. Esse método é aplicado em um conjunto de imagens de microscopia por fluorescência de 32 canais, em um experimento para identificar variações na estrutura foliar do espécime Jacaranda Caroba submetidos à diferentes condições. O método RDSA obtém as maiores taxas de acerto de classificação nesse conjunto de dados, com 91, 9% usando o esquema de validação cruzada Leave-one-out e 90, 5(±1, 1)% com 10-pastas, contra 81, 8% e 84, 7(±2, 2) da rede convolucional VGG16. / Multispectral images are present in the vast majority of current imaging devices, from personal cameras to microscopes, telescopes and satellites. However, much of the work in texture analysis and the like proposes monochromatic approaches, which often consider only gray levels. In this context and considering the performance increase of current computers, the use of the spectral information must be considered in the construction of better models. Lately, pre-trained deep convolutional neural networks have been used in 3-channel color images, however they are limited to just this format and compute many convolutions, which demands specific hardware (GPU). These facts motivated this work, which propose techniques for the modeling and characterization of multispectral images based on complex networks, which has proved to be an efficient tool in previous works and has computational complexity similar to traditional methods. Two approaches are introduced for application in 3-channel color images, called Multilayer Network (RM) and Directed Multilayer Network (RMD). These methods model all channels of the image together, where the networks have intra- and inter-channel connections, similar to the opponent color processing of the human visual system. Experiments in five color texture datasets shows that the RMD proposal overcomes several methods of the literature in general, including convolutional networks and traditional integrative methods. In addition, the proposals have demonstrated high robustness to different color spaces (RGB, LAB, HSV and I1I2I3), while other methods oscillate from dataset to dataset. Moreover it is proposed a new method to characterize multispectral images of many channels, called Directed Network of Angular Similarity (RDSA). In this proposal, each multispectral pixel is considered as a vector of dimensions equivalent to the number of channels of the image and the weight of the edges represents its cosine similarity, pointing to the pixel of greatest absolute value. This method is applied to a set of fluorescence microscopy images of 32 channels in an experiment to identify variations in the leaf structure of the Jacaranda Caroba specimen under different conditions. The RDSA method obtains the highest classification rates in this dataset, with 91.9% with the Leave-one-out cross-validation scheme and 90.5(±1.1)% with 10-folds, against 81.8% and 84.7(±2.2) of the convolutional network VGG16.

O aumento da expectativa de vida, associado a hábitos menos saudáveis da população, faz com que a probabilidade de desenvolvimento de algum tipo de tumor aumente. Por isso, o câncer tem se tornado cada vez mais um problema importante em saúde pública. Os métodos atualmente empregados para a detecção não são eficientes para um diagnóstico rápido e preciso, levando à busca de novas técnicas. A utilização de imagens ópticas têm se mostrado uma boa alternativa, para a realização de um diagnóstico precoce. Neste trabalho, apresentamos um sistema de imagem de campo amplo para detecção óptica de alterações teciduais, baseado em fluorescência e refletância. O desenvolvimento do sistema de imagem envolve tanto a montagem do protótipo como a proposta de um processamento das imagens adquiridas. Para a aquisição das imagens de fluorescência e refletância foi construído um sistema que utiliza uma câmera CCD colorida de alta resolução, juntamente com uma fonte de iluminação baseada em Diodos Emissores de Luz (LEDs). Para as imagens de fluorescência, utilizou-se um LED emitindo em 400 nm, juntamente com um filtro óptico, para obtenção apenas do sinal de fluorescência. As imagens de refletância foram feitas em cinco regiões: UV; azul; verde; vermelho e luz branca. Com a aquisição dessas imagens é possível formar uma imagem multiespectral da região analisada, sendo que cada tipo de imagem fornece uma informação diferente sobre o tecido analisado. Para a determinação de regiões que apresentam características ópticas distintas, utilizou-se o algoritmo k-means, que através do cálculo da distância geométrica entre as amostras, separa regiões opticamente distintas. Para a validação do sistema, foi utilizado um modelo in vivo, através da indução de lesões de pele por exposição a raios UV em camundongos hairless. Para formar a imagem multiespectral de uma lesão, foram adquiridas a imagem de fluorescência e as cinco imagens de refletância nas diferentes regiões. Para completar a imagem multiespectral, uma imagem da razão entre as componentes vermelha e verde da imagem de fluorescência foi adicionada, pois durante o desenvolvimento de uma lesão neoplásica, há uma alteração nessa proporção. A utilização de diferentes tipos de imagem permite um aumento do contraste na discriminação entre diferentes regiões. Através da utilização de fluorescência e refletância para a formação de imagens multiespectrais e de um processamento de imagens, foi possível delimitar áreas opticamente diferentes, resultado importante para a detecção e delineamento da lesão. / The increase of life expectancy associated to healthless habits results in a higher probability of tumor development. In this sense, cancer has been increasingly considered an important health public concern. The conventional methods for cancer detection are not efficient for the acquisition of a fast and precise diagnostics. In this study we present a widefield imaging system for optical detection of tissue changes based on fluorescence and reflectance. The development of the imaging system involves the device assembly as well as a proposed image processing. For the acquisition of fluorescence and reflectance images a system with a high resolution color CCD camera together with LED-based light source was built. A LED emitting at 400 nm and an optical filter were used for fluorescence imaging. The reflectance images were acquired at five spectral intervals: UV, blue, green, red, and white. After the acquisition of this set of images it is possible to merge a multispectral image of the target tissue, where each image type provides distinct information from the investigated sample. For the determination of the regions that present different optical characteristics, a k-means algorithm was used. An in vivo animal model of UV-induced skin lesions at hairless mice was used for system validation. In order to obtain the multispectral image of a lesion, a fluorescence image and five reflectance images were acquired. To complete the multispectral image, an image of the ratio of the red to green components of fluorescence image was added, because during malignant development a change of this ratio is observed. The use of different image types allows the increase of the discrimination contrast of distinct regions after an image processing it was possible to discriminate optically different regions, a result that is relevant for lesion detection and delimitation.

This paper evaluates the classification accuracy of three neural network classifiers on a satellite image-based pattern classification problem. The neural network classifiers used include two types of the Multi-Layer-Perceptron (MLP) and the Radial Basis Function Network. A normal (conventional) classifier is used as a benchmark to evaluate the performance of neural network classifiers. The satellite image consists of 2,460 pixels selected from a section (270 x 360) of a Landsat-5 TM scene from the city of Vienna and its northern surroundings. In addition to evaluation of classification accuracy, the neural classifiers are analysed for generalization capability and stability of results. Best overall results (in terms of accuracy and convergence time) are provided by the MLP-1 classifier with weight elimination. It has a small number of parameters and requires no problem-specific system of initial weight values. Its in-sample classification error is 7.87% and its out-of-sample classification error is 10.24% for the problem at hand. Four classes of simulations serve to illustrate the properties of the classifier in general and the stability of the result with respect to control parameters, and on the training time, the gradient descent control term, initial parameter conditions, and different training and testing sets. (authors' abstract) / Series: Discussion Papers of the Institute for Economic Geography and GIScience

O aumento da expectativa de vida, associado a hábitos menos saudáveis da população, faz com que a probabilidade de desenvolvimento de algum tipo de tumor aumente. Por isso, o câncer tem se tornado cada vez mais um problema importante em saúde pública. Os métodos atualmente empregados para a detecção não são eficientes para um diagnóstico rápido e preciso, levando à busca de novas técnicas. A utilização de imagens ópticas têm se mostrado uma boa alternativa, para a realização de um diagnóstico precoce. Neste trabalho, apresentamos um sistema de imagem de campo amplo para detecção óptica de alterações teciduais, baseado em fluorescência e refletância. O desenvolvimento do sistema de imagem envolve tanto a montagem do protótipo como a proposta de um processamento das imagens adquiridas. Para a aquisição das imagens de fluorescência e refletância foi construído um sistema que utiliza uma câmera CCD colorida de alta resolução, juntamente com uma fonte de iluminação baseada em Diodos Emissores de Luz (LEDs). Para as imagens de fluorescência, utilizou-se um LED emitindo em 400 nm, juntamente com um filtro óptico, para obtenção apenas do sinal de fluorescência. As imagens de refletância foram feitas em cinco regiões: UV; azul; verde; vermelho e luz branca. Com a aquisição dessas imagens é possível formar uma imagem multiespectral da região analisada, sendo que cada tipo de imagem fornece uma informação diferente sobre o tecido analisado. Para a determinação de regiões que apresentam características ópticas distintas, utilizou-se o algoritmo k-means, que através do cálculo da distância geométrica entre as amostras, separa regiões opticamente distintas. Para a validação do sistema, foi utilizado um modelo in vivo, através da indução de lesões de pele por exposição a raios UV em camundongos hairless. Para formar a imagem multiespectral de uma lesão, foram adquiridas a imagem de fluorescência e as cinco imagens de refletância nas diferentes regiões. Para completar a imagem multiespectral, uma imagem da razão entre as componentes vermelha e verde da imagem de fluorescência foi adicionada, pois durante o desenvolvimento de uma lesão neoplásica, há uma alteração nessa proporção. A utilização de diferentes tipos de imagem permite um aumento do contraste na discriminação entre diferentes regiões. Através da utilização de fluorescência e refletância para a formação de imagens multiespectrais e de um processamento de imagens, foi possível delimitar áreas opticamente diferentes, resultado importante para a detecção e delineamento da lesão. / The increase of life expectancy associated to healthless habits results in a higher probability of tumor development. In this sense, cancer has been increasingly considered an important health public concern. The conventional methods for cancer detection are not efficient for the acquisition of a fast and precise diagnostics. In this study we present a widefield imaging system for optical detection of tissue changes based on fluorescence and reflectance. The development of the imaging system involves the device assembly as well as a proposed image processing. For the acquisition of fluorescence and reflectance images a system with a high resolution color CCD camera together with LED-based light source was built. A LED emitting at 400 nm and an optical filter were used for fluorescence imaging. The reflectance images were acquired at five spectral intervals: UV, blue, green, red, and white. After the acquisition of this set of images it is possible to merge a multispectral image of the target tissue, where each image type provides distinct information from the investigated sample. For the determination of the regions that present different optical characteristics, a k-means algorithm was used. An in vivo animal model of UV-induced skin lesions at hairless mice was used for system validation. In order to obtain the multispectral image of a lesion, a fluorescence image and five reflectance images were acquired. To complete the multispectral image, an image of the ratio of the red to green components of fluorescence image was added, because during malignant development a change of this ratio is observed. The use of different image types allows the increase of the discrimination contrast of distinct regions after an image processing it was possible to discriminate optically different regions, a result that is relevant for lesion detection and delimitation.

Imagens multiespectrais estão presentes na grande maioria de dispositivos de imageamento atuais, desde câmeras pessoais até microscópios, telescópios e satélites. No entanto, grande parte dos trabalhos em análise de texturas e afins propõem abordagens monocromáticas, que muitas vezes consideram apenas níveis de cinza. Nesse contexto e considerando o aumento da capacidade dos computadores atuais, o uso da informação espectral deve ser considerada na construção de modelos melhores. Ultimamente redes neurais convolucionais profundas pré-treinadas tem sido usadas em imagens coloridas de 3 canais, porém são limitadas a apenas esse formato e computam muitas convoluções, o que demanda por hardware específico (GPU). Esses fatos motivaram esse trabalho, que propõem técnicas para a modelagem e caracterização de imagens multiespectrais baseadas em redes complexas, que tem se mostrado uma ferramenta eficiente em trabalhos anteriores e possui complexidade computacional similar à métodos tradicionais. São introduzidas duas abordagens para aplicação em imagens coloridas de três canais, denominadas Rede Multicamada (RM) e Rede Multicamada Direcionada (RMD). Esses métodos modelam todos os canais da imagem de forma conjunta, onde as redes possuem conexões intra e entre canais, de forma parecida ao processamento oponente de cor do sistema visual humano. Experimentos em cinco bases de textura colorida mostram a proposta RMD supera vários métodos da literatura no geral, incluindo redes convolucionais e métodos tradicionais integrativos. Além disso, as propostas demonstraram alta robustez a diferentes espaços de cor (RGB, LAB, HSV e I1I2I3), enquanto que outros métodos oscilam de base para base. Também é proposto um método para caracterizar imagens multiespectrais de muitos canais, denominado Rede Direcionada de Similaridade Angular (RDSA). Nessa proposta, cada pixel multiespectral é considerado como um vetor de dimensão equivalente à quantidade de canais da imagem e o peso das arestas representa sua similaridade do cosseno, apontando para o pixel de maior valor absoluto. Esse método é aplicado em um conjunto de imagens de microscopia por fluorescência de 32 canais, em um experimento para identificar variações na estrutura foliar do espécime Jacaranda Caroba submetidos à diferentes condições. O método RDSA obtém as maiores taxas de acerto de classificação nesse conjunto de dados, com 91, 9% usando o esquema de validação cruzada Leave-one-out e 90, 5(±1, 1)% com 10-pastas, contra 81, 8% e 84, 7(±2, 2) da rede convolucional VGG16. / Multispectral images are present in the vast majority of current imaging devices, from personal cameras to microscopes, telescopes and satellites. However, much of the work in texture analysis and the like proposes monochromatic approaches, which often consider only gray levels. In this context and considering the performance increase of current computers, the use of the spectral information must be considered in the construction of better models. Lately, pre-trained deep convolutional neural networks have been used in 3-channel color images, however they are limited to just this format and compute many convolutions, which demands specific hardware (GPU). These facts motivated this work, which propose techniques for the modeling and characterization of multispectral images based on complex networks, which has proved to be an efficient tool in previous works and has computational complexity similar to traditional methods. Two approaches are introduced for application in 3-channel color images, called Multilayer Network (RM) and Directed Multilayer Network (RMD). These methods model all channels of the image together, where the networks have intra- and inter-channel connections, similar to the opponent color processing of the human visual system. Experiments in five color texture datasets shows that the RMD proposal overcomes several methods of the literature in general, including convolutional networks and traditional integrative methods. In addition, the proposals have demonstrated high robustness to different color spaces (RGB, LAB, HSV and I1I2I3), while other methods oscillate from dataset to dataset. Moreover it is proposed a new method to characterize multispectral images of many channels, called Directed Network of Angular Similarity (RDSA). In this proposal, each multispectral pixel is considered as a vector of dimensions equivalent to the number of channels of the image and the weight of the edges represents its cosine similarity, pointing to the pixel of greatest absolute value. This method is applied to a set of fluorescence microscopy images of 32 channels in an experiment to identify variations in the leaf structure of the Jacaranda Caroba specimen under different conditions. The RDSA method obtains the highest classification rates in this dataset, with 91.9% with the Leave-one-out cross-validation scheme and 90.5(±1.1)% with 10-folds, against 81.8% and 84.7(±2.2) of the convolutional network VGG16.

M.Ing. / Globally, hazardous weather phenomena such as violent storms, oods, cyclones, tornadoes, snow and hail contribute to signi cant annual xed property damages, loss of movable property and loss of life. The majority of global natural disasters are related to hydro-meteorological events. Hazardous storms are destructive and pose a threat to life and property. Forecasting, monitoring and detecting hazardous storms are complex and demanding tasks, that are however essential. In this study automatic hazardous weather detection utilizing remotely sensed meteorological data has been investigated. Image processing techniques have been analyzed and applied to multispectral meteorological satellite image data obtained from the Spinning Enhanced Visible and Infrared Imager (SEVIRI) instruments on-board the Meteosat Second Generation (MSG) geostationary meteorological satellites Meteosat-8 and Meteosat-9. The primary focus of this study is the detection of potentially hazardous hydrometeorological phenomena in South Africa. A methodology for detecting potentially hazardous storms over South Africa using meteorological satellite imagery from MSG/SEVIRI is presented. An index indicative of the hazardous potential of a storm is de ned to aid in the identi cation of a ected geographical areas and to quantify the destructive potential of the detected storm. The Hazardous Potential Index (HPI) is generated through the use of image processing techniques such as cloud masking, cloud tracking and an image-based analysis of the constituent elements of a severe convective storm. A retrospective review was performed with respect to 20 case studies of documented storms which had adversely a ected areas of South Africa. A red-green-blue (RGB) composite image analysis technique, that may be utilized in the identi cation of severe convective storms using SEVIRI image data, was also applied to these case studies.

No description available.

La recherche en biométrie a connu une grande évolution durant les dernières annéessurtout avec le développement des méthodes de décomposition de visage. Cependant,ces méthodes ne sont pas robustes particulièrement dans les environnements incontrôlés.Pour faire face à ce problème, l'imagerie multispectrale s'est présentée comme une nouvelletechnologie qui peut être utilisée en biométrie basée sur la reconnaissance de visage.Dans tous ce processus, la qualité des images est un facteur majeur pour concevoirun système de reconnaissance fiable. Il est essentiel de se disposer d'images de hautequalité. Ainsi, il est indispensable de développer des algorithmes et des méthodes pourl'amélioration des données sensorielles. Cette amélioration inclut plusieurs tâches tellesque la déconvolution des images, le defloutage, la segmentation, le débruitage. . . Dansle cadre de cette thèse, nous étudions particulièrement la suppression de bruit ainsi quela segmentation de visage.En général, le bruit est inévitable dans toutes applications et son élimination doit sefaire tout en assurant l'intégrité de l'information confinée dans l'image. Cette exigenceest essentielle dans la conception d'un algorithme de débruitage. Le filtre Gaussienanisotropique est conçu spécifiquement pour répondre à cette caractéristique. Nous proposonsd'étendre ce filtre au cas vectoriel où les données en disposition ne sont plus desvaleurs de pixels mais un ensemble de vecteurs dont les attribues sont la réflectance dansune longueur d'onde spécifique. En outre, nous étendons aussi le filtre de la moyennenon-local (NLM) dans le cas vectoriel. La particularité de ce genre de filtre est la robustesseface au bruit Gaussien.La deuxième tâche dans le but d'amélioration de données sensorielles est la segmentation.Le clustering est l'une des techniques souvent utilisées pour la segmentation etclassification des images. L'analyse du clustering implique le développement de nouveauxalgorithmes particulièrement ceux qui sont basés sur la méthode partitionnelle.Avec cette approche, le nombre de clusters doit être connu d'avance, chose qui n'est pastoujours vraie surtout si nous disposons de données ayant des caractéristiques inconnues.Dans le cadre de cette thèse, nous proposons de nouveaux indices de validationde clusters qui sont capables de prévoir le vrai nombre de clusters même dans le cas dedonnées complexes.A travers ces deux tâches, des expériences sur des images couleurs et multispectrales sontréalisées. Nous avons utilisé des bases de données d'image très connues pour analyserl'approche proposée. / In this thesis, we focus on multispectral image for face recognition. With such application,the quality of the image is an important factor that affects the accuracy of therecognition. However, the sensory data are in general corrupted by noise. Thus, wepropose several denoising algorithms that are able to ensure a good tradeoff betweennoise removal and details preservation. Furthermore, characterizing regions and detailsof the face can improve recognition. We focus also in this thesis on multispectral imagesegmentation particularly clustering techniques and cluster analysis. The effectiveness ofthe proposed algorithms is illustrated by comparing them with state-of-the-art methodsusing both simulated and real multispectral data sets.

Cette thèse propose une nouvelle méthode d’optimisation bio-inspirée basée sur le GWO avec pour but de pouvoir résoudre des problèmes d’optimisation dits mixtes, c’est-à-dire des problèmes composés de variables continues et discrètes. Cette nouvelle méthode baptisée mixed GWO est ensuite appliquée à 2 problématiques distincts.Tout d’abord, le mixed GWO pourra permettre d’améliorer la qualité de la classification d’image par SVM. En effet, la fiabilité d’un SVM va dépendre de ses paramètres d’entraînement, et il n’existe pas de méthode non empirique et non exhaustive permettant de définir ces paramètres pour un problème de classification donné. Le mixed GWO se propose comme une solution à ce problème de paramétrage. La classification doit permettre à l’entreprise IntuiSense d’ajouter une brique de reconnaissance de genre à son outil de mesure d’audience ISAM.Ensuite, le mixedGWO est employé pour faire du débruitage et du démêlage de spectres en simultanée sur des images multi-spectrales ou hyper-spectrales. En effet, la qualité du démêlage des spectres va être particulièrement dépendant de la qualité du débruitage de l’image : faire ces 2 étapes simultanément permet donc un gain de temps et une fiabilité des résultats bien plus intéressants que les faire l’une après l’autre. / This thesis proposes a novel bio-inspired optimization method based on the GWOalgorithm, with the purpose of solving mixed optimization problems, i.e. problems with bothcontinuous and discrete variables. This novel method is named mixedGWO and is applied to2 distinct problematics.Firstly, the mixedGWO should permit to improve the quality of image classification bySVM. Indeed, a SVM accuracy will depend of its training parameters, and there is nonempirical and non exhaustive method to define these parameters for a given classificationproblem. Therefore, the mixedGWO can be used as a solution to this parametring problem.The improve classification should allow the company IntuiSense to add the gender recognition feature to its audience measurement tool ISAM.Secondly, the mixedGWO is used for joint denoising and unmixing of spectra in multispectral and hyper-spectral image processing. Indeed, the unmixing’s quality is stronglydependent of the denoising quality : doing these 2 steps simultaneously permits a gain oftime and a results’ accuracy way better than if they are done one after the other.