Abstract Wastewater treatment is facing challenges due to the stricter legislation and increasing variety in the quantity and quality of the influent. In addition, for economic reasons the use of resources should be efficient. Therefore new automated monitoring tools and methods along with intelligent use of the gathered data are required to increase knowledge on the treatment process and to receive predictive information on the quality of the effluent that can be used to assist in optimizing the process operation. Predicting the effluent quality is difficult, as the complex treatment process includes several simultaneous nonlinear mechanisms and the relevant continuous information on the floc morphology is commonly missing, even though the flocculation process holds the key role in efficient wastewater treatment. Automated optical monitoring is able to reveal new valuable information on wastewater continuously, fast, objectively, and with minimal labor contribution. The main aim of this research was to develop predictive models for the quality parameters of treated wastewater at two full-scale wastewater treatment processes utilizing optical monitoring results. The actual predictive information was enabled using the measurements from the beginning of the process in model development alone. The optimal subsets of the model variables were sought by variable selection methods. The research also shows how utilization of the process measurements and the optical monitoring variables separately and together affect the modeling accuracy. The quality of the effluent in municipal and industrial wastewater treatment processes can be predicted in varying operating conditions by utilizing optical monitoring and process measurements. The prediction accuracy is sufficient to reveal the level of and changes in effluent quality. Thus, process operation can be optimized, more efficiently purified wastewater can be achieved, and environmental damage, health-related risks, and economic losses can be minimized. / Tiivistelmä Jätevedenkäsittely kohtaa tulevaisuudessa haasteita tiukentuvan lainsäädännön sekä lisääntyvän jätevesimäärän ja laatuvaihteluiden vuoksi. Lisäksi taloudellisista syistä resurssien käytön pitäisi olla mahdollisimman tehokasta. Tämän vuoksi tarvitaan uusia automaattisia monitorointivälineitä ja -menetelmiä sekä kerätyn datan älykästä käyttöä lisäämään tietoa käsittelyprosessista ja ennustamaan käsitellyn jäteveden laatua prosessin ohjauksen optimoimiseksi. Käsitellyn jäteveden laadun ennustaminen on vaikeaa, sillä monimutkainen käsittelyprosessi sisältää useita yhtäaikaisia epälineaarisia mekanismeja ja olennainen jatkuva-aikainen tieto flokkien morfologiasta puuttuu, vaikka flokkausprosessi on keskeisessä asemassa jätevedenpuhdistuksen tehokkaan toiminnan kannalta. Automaattisen optisen monitoroinnin avulla saadaan uutta hyödyllistä tietoa jätevedestä jatkuva-aikaisesti, nopeasti, objektiivisesti ja vähäisellä työpanoksella. Tämän tutkimustyön päätavoite oli kehittää ennustemalleja kahden täysikokoisen jätevesilaitoksen käsitellyn jäteveden laatuparametreille hyödyntäen automaattista optista monitorointia. Malleissa käytettiin vain prosessin alkuosan mittauksia, jotta malleilla olisi todellinen kyky ennakoida käsitellyn jäteveden laatu. Mallien optimaaliset tulomuuttujat etsittiin muuttujavalinta-algoritmeilla. Tutkimus näyttää myös, miten prosessimittausten ja optisen monitoroinnin muuttujien hyödyntäminen yhdessä ja erikseen vaikuttaa mallien tarkkuuteen. Optisen monitoroinnin hyödyntäminen yhdessä prosessimittausten kanssa mahdollistaa käsitellyn jäteveden laadun ennustamisen vaihtelevissa olosuhteissa sekä teollisessa että kunnallisessa jätevedenkäsittelylaitoksessa. Mallinnustarkkuus on riittävä laatuparametrien tason ja muutosten esittämiseen ja käytettäväksi apuna prosessin ohjauksessa. Tällöin jäteveden puhdistusprosessi tehostuu ja ympäristövahingot, terveysriskit ja taloudelliset menetykset voidaan minimoida.

Purpose: Automatic border detection is the first and most crucial step for lesion segmentation and can be very challenging, due to several lesion characteristics. There are many melanoma border-detecting algorithms that perform poorly on dermoscopy images of basal cell carcinoma (BCC), which is the most common skin cancer. One of the reasons for poor lesion detection performance is that there are very few algorithms that detect BCC borders, because they are difficult to segment, even for dermatologists. This difficulty is due to low contrast, variation in lesion color and artifacts inside/outside the lesion. Segmentation that has adequate lesion-feature capture, with acceptable tolerance, will facilitate accurate feature segmentation, thereby maximizing classification accuracy. Methods: The main objective of this research was to develop an effective BCC border detecting algorithm whose accuracy is better than the existing melanoma border detectors that have been applied to BCCs. Fifteen auto-thresholding techniques were implemented for BCC lesion segmentation; but, only five were selected for use in algorithm development. A novel technique was developed to automatically expand BCC lesion borders, to completely circumscribe the lesion. Two error metrics were used that better measure Type II (false-negative) errors: Relative XOR error and Lesion Capture Ratio (a novel error metric). Results: On training and test sets of 1023 and 119 images, respectively, based on two error metrics, five thresholding-based algorithms outperformed two state-of-the-art melanoma segmentation techniques, in segmenting BCCs. Five algorithms generated borders that appreciably better matched dermatologists’ hand-drawn borders which were used as the “gold standard.” Conclusion: The five developed algorithms, which included solutions for image-vignetting correction and border expansion, to achieve dermatologist-like borders, provided more inclusive and therefore, feature-preserving border detection, favoring better BCC classification accuracy, for future work.

In response to recent declines, forested riparian wetland areas have become an increased conservation and management area of concern focusing on increasing biodiversity and promoting healthy ecosystem services. Additionally, passerine birds have also experienced a sharp global decline in that associated habitat. To mitigate further declines of both habitat and species numbers government programs and agencies have intensified conservation efforts. However, the practices employed are often assumed to be beneficial without conducting dedicated surveys to measure efficacy and practicality of current approaches. As such, visual evidence and statistics are often needed to promote or validate further support and funding for continuing with current polices or creating new focal areas and practices. This study strives to provide an inexpensive, efficient way to assess conservation areas based on a target species through a generalized and adaptive methodology. The Cypress Creek National Wildlife Refuge in southern Illinois provides an opportunity to do just that with a focus on songbirds. The methodology outlined in this study implements multiple remote sensing land use and land cover classification techniques utilizing Landsat imagery from 1984 to 2014 to create a temporal analysis of the region from pre-refuge era to current refuge designated era. Field surveys from the 2015 songbird summer breeding and fall migration seasons as well as vegetation surveys for field-truthing supplement the remote sensing results. The classification methodology incudes a combination of pan-sharpening Landsat images to a 15 m x 15 m spatial resolution, texture analysis, object based image analysis, and Random Forests to produce land use and land cover maps. For the sake of comparison the same classification process is performed with the untransformed, source images at 30 m x 30 m spatial resolution. Landscape metrics such as the interspersion and juxtaposition index and the contiguity index also provide further insight to temporal landscape patterns. At the completion of the study it was found that there was a minimal difference between the overall classification accuracy of transformed and untransformed images and that lowest overall accuracy in the study was 91% while the highest was 98%. The key survey statistics concluded that during the summer and fall observation periods songbirds in forested wetland areas had a propensity to utilize areas closest to the wetland edge as opposed to inland areas. Furthermore, during fall migration it was concluded that the mixed forest habitat type had a direct effect on observation numbers. Overall, with the aid of multiple landscape metrics, it was shown that the region was increasing in forested area, patch density, and contiguity; in response the passerines were using the area at a high rate, especially near wetland edges creating a sustainable focal area for conservation and management. The methodology and results in this study contribute to an ongoing effort to provide visual and statistical evidence that is reliable and accessible for policy making. The potential to manipulate the generalized methods used in this study to enhance any land use and land class classifications and apply to any targeted species certainly exists. Future studies will want to investigate the use of higher spatial resolution images or actively take reflectance recordings in the field and supplement the temporal maps with a multi-year dedicated species dataset for maximum benefit.

Em um tratamento de águas de abastecimento ou residuárias, a coagulação/floculação é um processo de grande importância. A determinação dos agentes coagulante/floculante e a quantidade necessária de cada um deles são problemas rotineiros para os profissionais e podem influenciar na qualidade do tratamento. Determinar o tamanho dos flocos é fundamental para que a operação de floculação ocorra com melhor eficiência. Ainda, variáveis hidrodinâmicas, como a velocidade, volume das partículas e do recipiente, também têm a mesma importância nesta finalidade. No presente trabalho, a modelagem do processo de floculação via balanço populacional foi estudada com o objetivo de prever a formação de agregados/flocos em relação com o tempo até atingir o estado estacionário e em condições de operação distintas. Assim, um modelo de balanço populacional pela técnica de discretização por pivô fixo foi implementado no software EMSO. Um estudo foi realizado através de processamento/análise de imagens capturadas em um experimento de coagulação/floculação. As imagens obtidas do processo de floculação foram analisadas e foi verificado a distribuição do tamanho dos flocos. Com os dados de volume e concentração inicial de partículas foi possível a simulação no modelo matemático. Assim tornou-se possível observar o crescimento, a estabilização, o comportamento dos flocos em comparação com o modelo matemático e a frequência desses flocos em cada tamanho em diferentes condições, como o gradiente de velocidade e a concentração inicial de partículas. Os resultados obtidos em comparação com os dados experimentais e a simulação do modelo matemático foram possíveis verificar comportamento semelhante de distribuição de tamanho de flocos. A partir dessa análise, o modelo matemático mostrou-se satisfatório em prever o comportamento experimental. Em diferentes condições pode-se verificar que o gradiente de velocidade e a concentração inicial de partículas influenciam na formação e estabilização dos flocos. Este estudo teve a utilização de softwares de livre acesso para trazer a proposta de implementação a fim de trazer economia de agentes químicos utilizados e melhoria nos processos de operação da estação tratamento de efluentes. / In a wastewater treatment, coagulation / flocculation is a process of great importance. The determination of coagulant / flocculant agent and the required amount of each problems are routine for professional and can influence the quality of treatment. Determination the flocs size is essential for flocculation operation to occur with greater efficiency. Still, hydrodynamic variables, such as speed, particles volume and the container also have the same importance in this purpose. In the present work, modeling of flocculation process by population balance was studied in order to predict the formation of aggregates / flocs in relation to time to reach steady state and at different operating conditions. Thus, a population balance model by discretization technique for fixed pivot was implemented in EMSO software. A study was conducted through the processing / analysis of images captured on an experiment coagulation / flocculation. Images obtained in the flocculation process were analyzed and it was found the size distribution of flocs. With the volume and initial particle concentration values, simulation was possible in mathematical model. Thus it became possible to observe the growth, establization and behavior of flocs in comparison with the mathematical model, and the frequency of such flocs in each size in different conditions, such as the velocity gradient and the initial particle concentration. The results obtained in comparison to experimental data and simulation of the mathematical model we observed behavior flocs size distribution similar. From this analysis, mathematical model was satisfactory to predict the experimental behavior. Under different conditions can be seen that the velocity gradient and the initial concentration of particles influence the formation and stabilization of flocs. This study was the use of software-free access to bring the implementation of proposal in order to bring savings of chemicals used and improvement in operating processes of wastewater treatment station.

O ciberespaço é tratado por alguns autores como um dos vetores de desterritorialização, problemática que assume importância nas ciências que tratam do espaço na modernidade tardia. Por um lado, por provocar uma evasão da materialidade; por outro, em uma perspectiva que interessa mais a este trabalho, por se configurar como um ambiente de enunciações mais livres do que as possíveis antes de seu advento, rompendo com discursos totalizantes acerca do espaço. Este trabalho examina como a cidade pode se construir em discursos no ambiente online. Estes discursos são tomados como territorializações, sucessão de linhas de des/reterritorialização que constituem um dizer território, e nestas construções, assim como na construção da cidade material, operam agentes táticos e estratégicos. Interessam em particular os discursos na forma imagética. Estes são entendidos como agenciamentos de diversos discursos e têm seu sentido produzido em uma relação dialógica com o contexto de enunciação. A dissertação apresenta um estudo de caso da construção discursiva em imagens do projeto Cais Mauá, em Porto Alegre, no site de relacionamentos Facebook. A pesquisa examinou doze imagens postadas em dois agrupamentos na plataforma. Para a coleta destas imagens, foi utilizada a ferramenta Netvizz. Para a análise destas imagens foi construído um método que incorporou descritores formais e icônicos das imagens a elementos de análise de discurso da hermenêutica dialógica. Com este método, se construiu o signo Cais Mauá para cada um dos agrupamentos e posteriormente o signo Cais Mauá que se produz na relação dialógica entre os dois agrupamentos. O trabalho contribui para a metodologia de análise visual e para a compreensão da cidade no contexto contemporâneo de aumento das relações mediadas pelo computador. / Some authors have held ciberspace as a vector of deterritorialization, a problem that has risen in importance in late modernity among sciences that deal with space. On the one hand because ciberspace incurs in an evasion of materiality; on the other hand, more relevant to this work, because it presents a space capable of more free enunciation than it was possible before its advent, rupturing with totalizing discourses about space. This work examines how the city can be constructed in discourses in this online environment. These discourses are held as territorializations, successions of de/reterritorializations that constitute a speech on territory. In these constructions, as in the construction of the material city, both strategic and tactical agents operate. Discourses in the form of images are particularly relevant. Held as collective agencies, their meaning is produced from the dialogical relation to the context of enunciation. This dissertation presents a case study on the imagetic discoursive construction of the Cais Mauá project, in Porto Alegre, on Facebook, a social media platform. The research analyzed twelve imagens posted to two groupings on the website. For data collection, the Netvizz tool was used. A method was built for the analysis of these images, to which formal and iconic descriptors were incorporated, as well as elements of dialogic hermeneutics, a method of discourse analysis. By this method, the sign Cais Mauá was constructed according to each of the groupings and later according to the dialogical relationship between the two groupings. This work contributes to visual analysis methodology and to the understanding of the city in a contemporary context of computer-mediated relationships.

Câmaras não métricas acopladas a Veículos Aéreos Não Tripulados (VANT) possibilitam coleta de imagens com alta resolução espacial e temporal. Além disso, o custo de operação e manutenção desses equipamentos são reduzidos. A classificação da cobertura da terra por meio dessas imagens são dificultadas devido à alta variabilidade espectral dos alvos e ao grande volume de dados gerados. Esses contratempos são contornados utilizando Análise de Imagens Baseada em Objetos (Object-Based Image Analysis – OBIA) e algoritmos de mineração de dados. Um algoritmo empregado na OBIA são as Árvores de Decisão (AD). Essa técnica possibilita tanto a seleção de atributos mais informativos quanto a classificação das regiões. Novas técnicas de AD foram desenvolvidas e, nessas inovações, foram inseridas funções para selecionar atributos e para melhorar a classificação. Um exemplo é o algoritmo C5.0, que possui uma função de redução de dados e uma de reforço. Nesse contexto, este trabalho tem como objetivo (i) avaliar o método de segmentação por crescimento de regiões em imagens com altíssima resolução espacial, (ii) determinar os atributos preditivos mais importantes na discriminação das classes e (iii) avaliar as classificações das regiões em relação aos parâmetros de seleção dos atributos (winnow) e de reforço (trial), que estão contidos no algoritmo C5.0. A segmentação da imagem foi efetuada no programa Spring, já as regiões geradas na segmentação foram classificadas pelo modelo de AD C5.0, que está disponível no programa R. Como resultado foi identificado que a segmentação crescimento de regiões possibilitou uma alta correspondência com regiões geradas pelo especialista, resultando em valores de Reference Bounded Segments Booster (RBSB) próximos a 0. Os atributos mais importantes na construção dos modelos por AD foram a razão entre a banda do verde com a azul (r_v_a) e o Modelo Digital de Elevação (MDE). Para o parâmetro de reforço (trial), não foi identificada melhora na acurácia da classificação ao aumentar seu valor. Já o parâmetro winnow possibilitou uma redução no número de atributos preditivos, sem perdas estatisticamente significativas na acurácia da classificação. A função de reforço (trial) não melhorou a classificação da cobertura da terra. Também não foram constatadas diferenças estatisticamente significativas quando winnow selecionado como verdadeiro, mas se encontrou o benefício desse último parâmetro reduzindo a dimensionalidade dos dados. Nesse sentido, este trabalho contribuiu para a classificação da cobertura da terra em imagens coletadas por VANT, uma vez que se desenvolveu algoritmos para automatizar os processos da OBIA e para avaliar a classificação das regiões em relação às funções de reforço (winnow) e de seleção do atributo (winnow) do classificador por árvore de decisão C5.0. / Non-metric cameras attached to Unmanned Aerial Vehicles (UAV) enable collection of images with high spatial and temporal resolution. In addition, the cost of operation and maintenance of equipment are reduced. The land cover classification through these images are hampered due to high spectral variability of the targets and the large volume of data generated. These setbacks are contoured using Image Analysis Based on Objects (OBIA) and data mining algorithms. An algorithm used in OBIA are Decision Trees (AD). This technique allows the selection of the most informative attributes as the classification of regions. New AD techniques have been developed and these innovations, were functions inserted to select attributes and to improve classification. One example is a C5.0 algorithm, which has a data reduction function and of boosting. In this context, this paper aims to (i) evaluate the segmentation method for growing regions in images with high spatial resolution, (ii) determine the most important predictive attributes in the discrimination of classes and (iii) evaluate the classifications of regions regarding the attributes selection parameters (winnow) and boosting (trial), which are contained in the C5.0 algorithm. The image segmentation was performed in Spring program, since the regions generated in segmentation were classified by model C5.0 , which is available in the program R. As a result it was identified that the segmentation by region growing provided a high correlation with regions generated by the expert, resulting in Reference Bounded Segments Booster values (RBSB) near 0. The most important features in the construction of models of decision tree are the ratio between the band of green with the blue (r_v_a) and the Digital Elevation Model (DEM). Was not identified improvement in classification accuracy when was increased value of trial parameter. Already winnow parameter enabled a reduction in the number of predictive attributes, with no statistically significant losses in the accuracy of the classification. The boosting function (trial) did not improve the classification of land cover. Also were not found statistically significant differences when winnow selected as true, but was found the benefit of the latter parameter to reducing the dimensionality of the data. Thus, this work contributed to the land cover classification in images collected by UAV, once that were developed algorithms to automate the processes of integration OBIA and decision tree (C5.0).

Curvas granulométricas de alimentação e produto em etapas de britagem permitem modelar e avaliar o desempenho dos equipamentos. Essa pesquisa foi baseada na determinação de curvas granulométricas tendo como ponto de partida estudos sobre conceitos de amostragem de Gy, Napier-Munn, Minnitt e Pittard, sendo que as amostragens para definição dessas curvas granulométricas utilizaram análises de imagens. A importância deste estudo está relacionada à necessidade de conhecer o minério que abastece a usina, pois nesse caso de estudo, com o aprofundamento da cava, este vem se tornando mais compacto e resistente. Um dos objetivos essenciais é utilizar a amostragem via análise de imagens como ferramenta para determinação da granulometria de ROM alimentado na britagem primária de uma planta e aplicar a técnica nas demais fases. Foram definidas duas metodologias, uma para amostragem do ROM encaminhado à britagem primária e outra para alimentação da britagem terciária. Inicialmente, na amostragem de ROM (itabiritos friáveis), foi identificada a bancada de trabalho para que uma carga desmontada (259t) fosse basculada. Houve segregação dos blocos maiores que 2,5cm para análise e mensuração por imagens e formação da pilha para quarteamento e peneiramento. Os dados foram tratados e analisados, gerando as curvas granulométricas. A metodologia para amostrar a granulometria da alimentação da britagem terciária (itabiritos friáveis e granulados) partiu da identificação do ponto de amostragem, seguida de filmagens e determinação das imagens a serem utilizadas, simulando um amostrador automático. Os dados foram analisados e tratados com auxílio estatístico para obtenção das curvas. As metodologias foram consideradas de grande valia pois forneceram dados de granulometria com reprodutibilidade, porém mostraram-se laboriosas. Com relação ao ROM, o protocolo amostral apresenta taxas de redução de massa elevadas. As curvas abrangem ampla faixa de tamanhos, indicando que, apesar das etapas de homogeneização e divisão, os erros de viés e precisão (relativa = ±1%) resultantes frente à massa do lote retrocalculado (259t) foram desprezíveis. Nas curvas de ROM, 65% do material está retido em APA, 4% está na faixa entre APA e APF e 5% está retido em 1m, mostrando que frente amostrada atende às especificações do equipamento. Com relação à britagem terciária, foram realizados dois testes. Cada imagem gerou uma curva e cada teste gerou quatro curvas que geraram uma curva média. A curva média foi comparada com a de projeto, sendo a curva de granulado a mais próxima. As maiores variações ocorreram nos maiores tamanhos, acima de 80mm para granulado e de 50mm para friável. Também, foi observada a mesma quantidade de material entre APA e APF, porém 100% do material granulado alimentado nesta etapa estava retido em APF, indicando impactos nesta operação. As análises das curvas permite a identificação de um cenário sobre mudanças na granulometria do ROM para os próximos anos. As metodologias foram desenvolvidas para proceder no futuro com a análise de materiais mais resistentes e se tornam de grande valia, pois a tendência do material é se tornar cada vez mais compacto. Desta forma, os resultados defendem este tipo de amostragem como uma técnica viável e sugerem sua utilização para controle de produtividade em outros depósitos. / Feed size distribution curves in crushing stages allows modeling and predict performance of crushers. This research was based on particle size’s distribution curves determination with start on studies about sampling theory from Gy, Napier-Munn, Minnitt e Pittard, however, the sampling was developed with image analysis techniques and not by conventional granulometry analysis (screening). The importance of this research is related to the necessity of knowing ore that feeds plant because, in this study, with increased pit ore is becoming stronger and resistant. One of the essential objective is to use sampling with image analysis techniques as a tool for determining ROM’s particle size that feeds the first stage of crushing plant and apply that techniques in other stages. It was defined two methodologies, one was developed to sample in situ the ROM granulometry that feeds the first crushing stage and the other one was developed specially to knows the particle size (granulometry) that feeds the third crushing stage. Initially, on sampling ROM (friable itabirite), it was identified the workbench to dump a mass of 259t. Big blocks were segregated. In order to measure big blocks were applied image analysis techniques. Smaller material than 2.5cm was screened. Data were processed and analyzed, generating the size distribution curves. The methodology to sampling the third crushing stage (friable itabirite and compact itabirite) began with the sampling point identification followed by recordings and images determination simulating an automatic sampler. Data were processed and analyzed with statistical help to obtain curves. Both methodologies were considered valuable as they provided size distribution curves with reproducibility but were laborious. ROM granulometry that was evaluated has shown insignificant bias and reproducibility errors. The sampling protocol shows high reduction of mass rates. ROM curves shows that 65% of material is retained on APA, 4% is retained in size range between APA and APF and 5% is retained in 1m, showing that front sampled are attending the equipments specifications. With respect to tertiary crushing, two tests were performed. Each image generated one curve and each test generated four curves that generated a medium curve. The medium curve was compared with project curve and the compact curve was the curve that stays closer. The largest errors are, as expects, to blocks of larger size, more than 80mm for compact and 50mm for friable. Also was observed the same material quantity on APA and APF but 100% of compact itabirite was retained in APF, indicating impacts in this operation. The curves analysis allows the identification about changes in ROM’s particle size for the coming years. The methodologies have been developed to proceed in the future with the analysis of more resistant materials and become of great value, because the tendency of the material is becoming increasingly compact. Thus, the results defends this kind of sampling as a viable technique and suggest their use for control of productivity of other mineral deposits.

Made available in DSpace on 2014-06-11T19:34:58Z (GMT). No. of bitstreams: 0 Previous issue date: 2009-10-19Bitstream added on 2014-06-13T19:04:56Z : No. of bitstreams: 1 goncalves_vd_dr_guara.pdf: 2777682 bytes, checksum: 5d4369fe1e30d71b1e79eccb5580ba09 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Este trabalho descreve o desenvolvimento de um sistema de análise de imagens de partículas de desgaste encontradas em óleos lubrificantes de equipamentos industriais. Para tanto, foi utilizado um sistema de aquisição de imagem para capturar imagens de amostras de óleo retidas em membranas de filtro e, também, desenvolvida uma metodologia analítica que faz a classificação quantitativa e qualitativa das partículas, relacionando-as ao modo de desgaste na qual foram geradas. Para a classificação quantitativa utilizou-se a norma ISO 4406 e para a qualitativa, a análise por meio de Redes Neurais Artificiais. O sistema aplicado consiste em uma câmera digital, um microscópio óptico monocular, um sistema de filtragem de óleo e dois programas computacionais desenvolvidos para realização da análise automatizada das imagens das partículas de desgaste obtidas. Foram utilizados óleos minerais de redutores de máquinas industriais no processo de obtenção de amostras. As partículas de desgaste foram analisadas pelo processo de microscopia óptica para obtenção da imagem da amostra relacionada ao desgaste. Os resultados obtidos mostraram que o sistema de análise desenvolvido realiza a classificação individual das partículas, através de Redes Neurais Artificiais, com uma eficiência de até 96%, além de analisar as múltiplas partículas contidas nas amostras, exibindo um relatório de acompanhamento e evolução do desgaste. Os programas computacionais desenvolvidos para essa análise possuem interface gráfica de fácil utilização. Eles podem ser amplamente utilizados no estudo e avaliação das partículas de desgaste obtidas de amostras de óleos industriais em empresas ou em universidades para fins educacionais. O sistema pode ser utilizado para análise de superfícies de metais obtidos pelo processo de metalografia ou na análise de imagens de microorganismos obtidas... / This paper describes the development of an image analysis system for wear particles found in industrial equipment lubricating oil. Hence, it was utilized an image acquisition system to capture image samples of the oil held in filter membranes. An analytical methodology was also developed to classify the particles quantitatively and qualitatively, relating them to the wear mode where they had been generated. The ISO 4406 standard was applied for quantitative classification and Artificial Neural Networks were used for qualitative classification analysis. The applied system consists of a digital camera, a monocular optical microscope, one oil filtering system and two software programs developed to perform the automated analysis of the acquired wear particles images. Mineral oils of gearbox industrial machines were used in the process of achieving the samples. The wear particles were analyzed by an optical microscopy system, to obtain the sample image related to the wear. The results showed that the analysis system that was developed classifies the individual particles through Artificial Neural Networks with 96% accuracy, in addition to analyzing the multiple particles contained in the samples and generating an evaluation report and wear evolution. The software programs developed for the analysis have a graphical interface easy to use. They can be used in the study and evaluation of the wear particles obtained from industrial oil samples in the companies or in the universities for educational purposes. The system can be applied to analyze the surface of metals acquired by metallographic processes or in the image analysis of microorganisms acquired from blood samples, opening a wide field of application within universities and researches.

[pt] O minério de ferro é um material policristalino oriundo de processos naturais complexos. Os minerais mais comuns que o compõem (hematita, magnetita, goethita, etc.) podem ser identificados no microscópio ótico de luz refletida, através de suas refletâncias distintas. A importância do estudo das hematitas, especificamente, surge porque os maiores depósitos de minério de ferro, no Brasil, são praticamente todos do tipo hematítico, com altos teores de ferro. A hematita é um mineral fortemente anisotrópico que apresenta pleocroísmo de reflexão. Isto faz com que o brilho na imagem mude com diferentes orientações dos cristais. Assim, quando se utiliza luz polarizada, o contraste entre os cristais aumenta o suficiente para diferenciá-los. Tradicionalmente, as hematitas são classificadas em tipos texturais identificados como hematita microcristalina (Mc), martita (Ma), e partículas policristalinas compactas (Co) formadas, por sua vez, de cristais dos tipos: granular (Gr), lamelar (La), lobular (Lo). Em trabalhos anteriores foi desenvolvida uma rotina de classificação automática para os diferentes tipos de hematitas. Esta rotina utiliza como entrada duas imagens de uma mesma região, a primeira em campo claro (CC) e a segunda polarizada circularmente (CPOL). Neste trabalho foram implementadas modificações nas etapas de aquisição das imagens CPOL e no processamento de ruído, visando melhorar as etapas de classificação. Assim, a imagem CPOL, que apresenta problemas característicos de iluminação, passou a ser capturada utilizando o recurso de sub-quadros (subframe), o que elimina a necessidade de correção de fundo, melhorando a qualidade dos mosaicos de imagens capturados. Em seguida, explorou-se recurso de saturação digital da câmera, de forma a melhorar, substancialmente, o contraste entre os cristais de hematita. Finalmente, testou-se o impacto do uso de um novo filtro de redução de ruído – Non-Local Means (MNL) – sobre a segmentação de cristais. Os resultados mostraram uma melhora substancial na identificação dos tipos texturais de hematita com relação aos métodos anteriores e também superiores à tradicional identificação visual por operador treinado. / [en] Iron ore is a polycrystalline material originated from complex natural processes. Its main composing minerals (hematite, magnetite, goethite, etc.) can be identified with the reflected light optical microscope through their distinctive reflectances. The relevance of studying hematite, specifically, originates from the fact that the largest Brazilian iron deposits are mostly of the hematitic type, with high iron content. Hematite is a strongly anisotropic mineral that presents reflectance pleocroism. Thus, different crystal orientations produce different brightness and, when using polarized light, the contrast between crystals is strong enough to allow their discrimination. Traditionally, hematites are classified in textural types identified as microcrystalline (Mc), martite (Ma) and compact polycristalline (Co), composed of granula (Gr), lamellar (La) and lobular (Lo) crystals. An automatic classification routine for hematite types was developed in previous works. This routine takes as input two images of the same region, one in Bright Field and the second in Circular Polarization (CPOL). In this work, modifications in the CPOL image acquisition and in noise filtering were implemented, in order to improve the classification step. Thus, the CPOL images, which present a characteristic background problem, were acquired employing the subframe method, what eliminates the need for background correction, improving the quality of image mosaics. Then, the digital saturation of the camera was optimized to improve substantially the contrast between hematite types. Finally, the impact of a new noise reduction filter – the Non-Local Means Filter – on crystal segmentation was evaluated. The results showed a substantial improvement in the identification of hematite textural types as compared to the previous method, and also superior to the traditional visual identification by an operator.

abstract: Accurate quantitative information of tumor/lesion volume plays a critical role in diagnosis and treatment assessment. The current clinical practice emphasizes on efficiency, but sacrifices accuracy (bias and precision). In the other hand, many computational algorithms focus on improving the accuracy, but are often time consuming and cumbersome to use. Not to mention that most of them lack validation studies on real clinical data. All of these hinder the translation of these advanced methods from benchside to bedside. In this dissertation, I present a user interactive image application to rapidly extract accurate quantitative information of abnormalities (tumor/lesion) from multi-spectral medical images, such as measuring brain tumor volume from MRI. This is enabled by a GPU level set method, an intelligent algorithm to learn image features from user inputs, and a simple and intuitive graphical user interface with 2D/3D visualization. In addition, a comprehensive workflow is presented to validate image quantitative methods for clinical studies. This application has been evaluated and validated in multiple cases, including quantifying healthy brain white matter volume from MRI and brain lesion volume from CT or MRI. The evaluation studies show that this application has been able to achieve comparable results to the state-of-the-art computer algorithms. More importantly, the retrospective validation study on measuring intracerebral hemorrhage volume from CT scans demonstrates that not only the measurement attributes are superior to the current practice method in terms of bias and precision but also it is achieved without a significant delay in acquisition time. In other words, it could be useful to the clinical trials and clinical practice, especially when intervention and prognostication rely upon accurate baseline lesion volume or upon detecting change in serial lesion volumetric measurements. Obviously, this application is useful to biomedical research areas which desire an accurate quantitative information of anatomies from medical images. In addition, the morphological information is retained also. This is useful to researches which require an accurate delineation of anatomic structures, such as surgery simulation and planning. / Dissertation/Thesis / Doctoral Dissertation Biomedical Informatics 2016