Marcelo Albuquerque - Academia.edu (original) (raw)

Papers by Marcelo Albuquerque

Applied Magnetic Resonance

Notas Técnicas

Resumo: A determinação da permeabilidadeé fundamental na caracterização de meios porosos em espec... more Resumo: A determinação da permeabilidadeé fundamental na caracterização de meios porosos em especial para estimar a capacidade de armazenamento e escoamento em formações particuladas. Diversos modelos baseados em tamanhos de grãos foram propostos para quantificar a permeabilidade. Todavia tais modelos apresentam limitações significativas que tornam seu uso restrito em diversas aplicações práticas. Neste trabalho apresentamos dois modelos, o primeiro, baseado no trabalho de Panda e Lake e o segundo, no trabalho de Kozeny e Carman, discutindo suas potenciais aplicações em imagens digitais de rochas geológicas e limitações com relaçãoà forma de medição de alguns parâmetros. Para analisar o desempenho destes modelos realizamos testes com imagens de microtomografia de amostras confeccionada com microesferas, buscando simular imagens de solo. As amostras sintéticas permitem construir um ambiente controlado para comparação com medidas obtidas apósà análise digital. Propomos uma cadeia de processamento de imagens para estimar medidas de permeabilidade e comparamos com quantidades obtidas através de simulações de fluxo. Verificamos que o modelo analítico Kozeny-Carman Generalizado obtivemos uma estimativa de permeabilidade absoluta em uma distribuição de partículas sintéticas com uma precisão de ±0.4µm 2 e ±61.1µm 2 (em duas amostras distintas). Com taxa de exatidão de 98% e 99% respectivamente, quando comparados com os valores obtidos através das CBPF-NT-009/17 11 simulações de fluxo.

Notas ténicas

A análise das imagens acústicas de poço tem como um de seus objetivos reconhecer e caracterizar f... more A análise das imagens acústicas de poço tem como um de seus objetivos reconhecer e caracterizar fraturas naturais. A interpretação dessas imagens em conjunto com diversos outros dados e análises, oferecem um panorama para determinar a sua viabilidade de exploração de petróleo e gás em um reservatório. A grande quantidade de dados de umúnico poço faz com que a análise visual consuma tempo relevante de profissionais dá area. Neste contexto, para auxiliar o desenvolvimento de métodos de reconhecimentos de padrões automatizados ou supervisionados para fraturas, avaliamos a performance de 5 métodos de segmentação: Gradiente Direcional, K-means, Otsu, Niblack e filtro Entropia Local. Para uma análise consistente, construímos um simulador de imagens acústicas com fraturas sintéticas representando diversos níveis de complexidade e varrendo o espaço de parâmetros do nosso modelo. Avaliamos a performance utilizando o percentual de sobreposição e correlação entre a segmentação e a região da fratura definida pela simulação. Em imagens sem ajuste de contraste para correção de iluminação o método K-means apresentou o melhor desempenho. No entanto, com utilização do ajuste de contraste os resultados com Otsu foram semelhantes ao K-means. O teste de hipótese unilateralà esquerda, para hipótese nula da média de correlação superiorà 50 %, a hipótese não foi rejeitada apenas pelos métodos K-means e Otsu, com valores respectivos de p ∼ 0.60 e p ∼ 0.61 dentro do intervalo σ.

Notas Técnicas

Formation permeability plays a crucial role in oil/gas industry as it helps geologists understand... more Formation permeability plays a crucial role in oil/gas industry as it helps geologists understand how the soil was formed and the most likely location of a certain reservoir. It also aids engineers and analysts to know whether that reservoir is exploitable or not. However, these measurements are very costly and difficult to obtain. In the past decades, several methods have appeared for estimating permeability from different sources like Nuclear Magnetic Resonance data or well-logs. In this work we tested the relationship between complex impedance spectra and permeability measurements, showing that-for the tested porous medium-it does exist a correlation of almost 98% between the obtained complex electrical impedance spectra and the permeability of those mediums. This link could be used in the oil/gas industry adapting the current electrical imaging tools to perform sweep measurements instead of single-frequency measurements, which could be used, in theory, to obtain permeability measurements with spacial resolution. On the other hand, we also designed and implemented an imaging tool that relies on this principle to obtain complex impedance volume images.

Proceedings of the 7 Simpósio Brasileiro de Geofísica

Notas Técnicas

Este documento apresenta um método de confecção de placas de circuito impresso (PCI) usando filme... more Este documento apresenta um método de confecção de placas de circuito impresso (PCI) usando filme fotográfico. O diferencial deste métodoé o uso reduzido de soluções químicas para se chegar ao resultado esperado. Enquanto outros métodos apresentam mais de duas soluções químicas, este utiliza apenas duas soluções para realizar toda a confecção de trilhas e ilhas em uma placa virgem de circuito impresso. Nos procedimentos são exploradas as propriedades de um filme fotográfico, o dryfilm para se desenhar ilhas e trilhas e assim alcançar o desenho projetado paro o circuito impresso. O métodoé uma proposta de baixo custo para fabricação de protótipos de média complexibilidade.

Journal of Petroleum Science and Engineering

Notas Técnicas, 2017

In this work we present a method to classify a set of rock textures based on a Spectral Analysis ... more In this work we present a method to classify a set of rock textures based on a Spectral Analysis and the extraction of the texture Features of the resulted images. Up to 520 features were tested using 4 different filters and all 31 different combinations were verified. The classification process relies on a Naïve Bayes classifier. We performed two kinds of optimizations: statistical optimization with covariance-based Principal Component Analysis (PCA) and a genetic optimization, for 10,000 randomly defined samples, achieving a final maximum classification success of 91% against the original ∼ 70% success ratio (without any optimization nor filters used). After the optimization 9 types of features emerged as most relevant.

Notas Técnicas, 2016

We present the Mediatrix filamentation method, a novell iterative procedure that decomposes elong... more We present the Mediatrix filamentation method, a novell iterative procedure that decomposes elongated objects in filaments along their main direction over their intensity peak. From this decomposition, the method measures the object's length and thickness. This technique is applied in preliminary tests to arc-shaped objects (simulated gravitational arcs) to recover their curvature center.

Notas Técnicas, 2016

In this work we discuss a method to classify a set of texturized images based on the extraction o... more In this work we discuss a method to classify a set of texturized images based on the extraction of their Haralick Features. This kind of Classification is capable of providing texture-based measurements (such as contrast or correlation) and use them as main parameters to classify the same type of patterns in other images. In order to improve the classification success ratio a spectral analysis of the textures and, therefore, the use of filters, before the classification step, is proposed here. In this work the classification success has been evaluated using Mean and Canny filters. On the other hand, the Principal Component Analysis is used to optimize the features extracted for the patterns on each image, before introduced into the classifier. With this method the classifying success ratio for the KTH-TIPS subset and 10,000 different permutations was increased-in averagefrom 72.28% to 84.25%.

Notas Técnicas, 2016

Resumo: A segmentação, realizada de maneira robusta, automatizada e eficiente, de diferentes fase... more Resumo: A segmentação, realizada de maneira robusta, automatizada e eficiente, de diferentes fases em imagens de microtomografia é um fator crítico e limitador na área de Petrofísica de Rocha Digital. Abordamos a questão partindo de um algoritmo com técnicas baseadas em filtros, obtendo a maximização da Entropia Local para definir um limiar entre fundo e objeto. Validamos a qualidade da segmentação a partir de imagens de amostras de microesferas de vidro, recuperamos o raio das esferas e comparamos a técnica proposta com outros dois algoritmos de segmentação.

We present the Mediatrix filamentation method, a novell iterative procedure that decomposes elong... more We present the Mediatrix filamentation method, a novell iterative procedure that decomposes elongated objects in filaments along their main direction over their intensity peak. From this decomposition, the method measures the object's length and thickness. This technique is applied in preliminary tests to arc-shaped objects (simulated gravitational arcs) to recover their curvature center.

Proceedings. XV Brazilian Symposium on Computer Graphics and Image Processing, 2002

This work aims at the implementation of Brazilian vehicle plate localization using image processi... more This work aims at the implementation of Brazilian vehicle plate localization using image processing techniques. In this paper, we show how to find the horizontal position of the plate based on its frequency characteristics in the whole vehicle image. Preliminary results show efficiencies above 96% for images in good illumination conditions.

Proceedings. XV Brazilian Symposium on Computer Graphics and Image Processing, 2002

In image processing, one of the most efficient techniques for image segmentation is entropy-based... more In image processing, one of the most efficient techniques for image segmentation is entropy-based thresholding. In this work a generalized entropy formalism that represents a recent development in statistical mechanics was applied. We propose, for the first time, an image thresholding method using a nonextensive entropy regarding the presence of nonadditive information content in image classes. Preliminary results are shown.

IEEE Latin America Transactions, 2008

Image analysis usually refers to processing of images with the goal of finding objects presented ... more Image analysis usually refers to processing of images with the goal of finding objects presented in the image. Image segmentation is one of the most critical tasks in automatic image analysis. The nonextensive entropy, also known as Tsallis entropy, is a recent development in statistical mechanics and has been considered as a useful measure in describing termostatistical properties of physical systems. In this new formalism a real quantity q was introduced as parameter for physical systems that presents long range interactions, long time memories and fractal-type structures. In image processing, one of the most efficient techniques for image segmentation is entropybased thresholding. This approach uses the Shannon entropy from the information theory considering the gray level image histogram as a probability distribution. In this work, it was applied the Tsallis entropy as a generalized entropy formalism for information theory. For the first time it was proposed an image thresholding method using a nonextensive relative entropy.

Scientometrics, 2009

In this work we have studied the research activity for countries of Europe, Latin America and Afr... more In this work we have studied the research activity for countries of Europe, Latin America and Africa for all sciences between 1945 and November 2008. All the data are captured from the Web of Science database during this period. The analysis of the experimental data shows that, within a nonextensive thermostatistical formalism, the Tsallis q-exponential distribution N (c) satisfactorily describes Institute of Scientific Information citations. The data which are examined in the present survey can be fitted successfully as a first approach by applying a single curve (namely, N (c) ∝ 1/[1 + (q − 1) c/T ] 1 q−1 with q ≃ 4/3 for all the available citations c, T being an "effective temperature". The present analysis ultimately suggests that the phenomenon might essentially be one and the same along the entire range of the citation number. Finally, this manuscript provides a new ranking index, via the "effective temperature" T , for the impact level of the research activity in these countries, taking into account the number of the publications and their citations.

Proceedings 13th Brazilian Symposium on Computer Graphics and Image Processing (Cat. No.PR00878), 2000

This work aims the recognition of characters in plates of private Brazilian vehicles using neural... more This work aims the recognition of characters in plates of private Brazilian vehicles using neural networks. Initially we define a standard for the image as prerequisite for the pre-processing. Afterward, the reduction of the dimensionality is made in order to make possible the use of the network. For this we establish a figure of merit-Q-characterizing, amongst some algorithms of characters splitting, the best one. Q is based on normalized cross correlation of the desired character image and the gotten one. The pre-processing allowed us the improvement of the image, reducing the background noise and increasing the contrast by applying a non-linear function. In the last stage, the neural network recognizes each character of the plate. Preliminary results are shown.

Applied Magnetic Resonance

Notas Técnicas

Resumo: A determinação da permeabilidadeé fundamental na caracterização de meios porosos em espec... more Resumo: A determinação da permeabilidadeé fundamental na caracterização de meios porosos em especial para estimar a capacidade de armazenamento e escoamento em formações particuladas. Diversos modelos baseados em tamanhos de grãos foram propostos para quantificar a permeabilidade. Todavia tais modelos apresentam limitações significativas que tornam seu uso restrito em diversas aplicações práticas. Neste trabalho apresentamos dois modelos, o primeiro, baseado no trabalho de Panda e Lake e o segundo, no trabalho de Kozeny e Carman, discutindo suas potenciais aplicações em imagens digitais de rochas geológicas e limitações com relaçãoà forma de medição de alguns parâmetros. Para analisar o desempenho destes modelos realizamos testes com imagens de microtomografia de amostras confeccionada com microesferas, buscando simular imagens de solo. As amostras sintéticas permitem construir um ambiente controlado para comparação com medidas obtidas apósà análise digital. Propomos uma cadeia de processamento de imagens para estimar medidas de permeabilidade e comparamos com quantidades obtidas através de simulações de fluxo. Verificamos que o modelo analítico Kozeny-Carman Generalizado obtivemos uma estimativa de permeabilidade absoluta em uma distribuição de partículas sintéticas com uma precisão de ±0.4µm 2 e ±61.1µm 2 (em duas amostras distintas). Com taxa de exatidão de 98% e 99% respectivamente, quando comparados com os valores obtidos através das CBPF-NT-009/17 11 simulações de fluxo.

Notas ténicas

A análise das imagens acústicas de poço tem como um de seus objetivos reconhecer e caracterizar f... more A análise das imagens acústicas de poço tem como um de seus objetivos reconhecer e caracterizar fraturas naturais. A interpretação dessas imagens em conjunto com diversos outros dados e análises, oferecem um panorama para determinar a sua viabilidade de exploração de petróleo e gás em um reservatório. A grande quantidade de dados de umúnico poço faz com que a análise visual consuma tempo relevante de profissionais dá area. Neste contexto, para auxiliar o desenvolvimento de métodos de reconhecimentos de padrões automatizados ou supervisionados para fraturas, avaliamos a performance de 5 métodos de segmentação: Gradiente Direcional, K-means, Otsu, Niblack e filtro Entropia Local. Para uma análise consistente, construímos um simulador de imagens acústicas com fraturas sintéticas representando diversos níveis de complexidade e varrendo o espaço de parâmetros do nosso modelo. Avaliamos a performance utilizando o percentual de sobreposição e correlação entre a segmentação e a região da fratura definida pela simulação. Em imagens sem ajuste de contraste para correção de iluminação o método K-means apresentou o melhor desempenho. No entanto, com utilização do ajuste de contraste os resultados com Otsu foram semelhantes ao K-means. O teste de hipótese unilateralà esquerda, para hipótese nula da média de correlação superiorà 50 %, a hipótese não foi rejeitada apenas pelos métodos K-means e Otsu, com valores respectivos de p ∼ 0.60 e p ∼ 0.61 dentro do intervalo σ.

Notas Técnicas

Formation permeability plays a crucial role in oil/gas industry as it helps geologists understand... more Formation permeability plays a crucial role in oil/gas industry as it helps geologists understand how the soil was formed and the most likely location of a certain reservoir. It also aids engineers and analysts to know whether that reservoir is exploitable or not. However, these measurements are very costly and difficult to obtain. In the past decades, several methods have appeared for estimating permeability from different sources like Nuclear Magnetic Resonance data or well-logs. In this work we tested the relationship between complex impedance spectra and permeability measurements, showing that-for the tested porous medium-it does exist a correlation of almost 98% between the obtained complex electrical impedance spectra and the permeability of those mediums. This link could be used in the oil/gas industry adapting the current electrical imaging tools to perform sweep measurements instead of single-frequency measurements, which could be used, in theory, to obtain permeability measurements with spacial resolution. On the other hand, we also designed and implemented an imaging tool that relies on this principle to obtain complex impedance volume images.

Proceedings of the 7 Simpósio Brasileiro de Geofísica

Notas Técnicas

Este documento apresenta um método de confecção de placas de circuito impresso (PCI) usando filme... more Este documento apresenta um método de confecção de placas de circuito impresso (PCI) usando filme fotográfico. O diferencial deste métodoé o uso reduzido de soluções químicas para se chegar ao resultado esperado. Enquanto outros métodos apresentam mais de duas soluções químicas, este utiliza apenas duas soluções para realizar toda a confecção de trilhas e ilhas em uma placa virgem de circuito impresso. Nos procedimentos são exploradas as propriedades de um filme fotográfico, o dryfilm para se desenhar ilhas e trilhas e assim alcançar o desenho projetado paro o circuito impresso. O métodoé uma proposta de baixo custo para fabricação de protótipos de média complexibilidade.

Journal of Petroleum Science and Engineering

Notas Técnicas, 2017

In this work we present a method to classify a set of rock textures based on a Spectral Analysis ... more In this work we present a method to classify a set of rock textures based on a Spectral Analysis and the extraction of the texture Features of the resulted images. Up to 520 features were tested using 4 different filters and all 31 different combinations were verified. The classification process relies on a Naïve Bayes classifier. We performed two kinds of optimizations: statistical optimization with covariance-based Principal Component Analysis (PCA) and a genetic optimization, for 10,000 randomly defined samples, achieving a final maximum classification success of 91% against the original ∼ 70% success ratio (without any optimization nor filters used). After the optimization 9 types of features emerged as most relevant.

Notas Técnicas, 2016

We present the Mediatrix filamentation method, a novell iterative procedure that decomposes elong... more We present the Mediatrix filamentation method, a novell iterative procedure that decomposes elongated objects in filaments along their main direction over their intensity peak. From this decomposition, the method measures the object's length and thickness. This technique is applied in preliminary tests to arc-shaped objects (simulated gravitational arcs) to recover their curvature center.

Notas Técnicas, 2016

In this work we discuss a method to classify a set of texturized images based on the extraction o... more In this work we discuss a method to classify a set of texturized images based on the extraction of their Haralick Features. This kind of Classification is capable of providing texture-based measurements (such as contrast or correlation) and use them as main parameters to classify the same type of patterns in other images. In order to improve the classification success ratio a spectral analysis of the textures and, therefore, the use of filters, before the classification step, is proposed here. In this work the classification success has been evaluated using Mean and Canny filters. On the other hand, the Principal Component Analysis is used to optimize the features extracted for the patterns on each image, before introduced into the classifier. With this method the classifying success ratio for the KTH-TIPS subset and 10,000 different permutations was increased-in averagefrom 72.28% to 84.25%.

Notas Técnicas, 2016

Resumo: A segmentação, realizada de maneira robusta, automatizada e eficiente, de diferentes fase... more Resumo: A segmentação, realizada de maneira robusta, automatizada e eficiente, de diferentes fases em imagens de microtomografia é um fator crítico e limitador na área de Petrofísica de Rocha Digital. Abordamos a questão partindo de um algoritmo com técnicas baseadas em filtros, obtendo a maximização da Entropia Local para definir um limiar entre fundo e objeto. Validamos a qualidade da segmentação a partir de imagens de amostras de microesferas de vidro, recuperamos o raio das esferas e comparamos a técnica proposta com outros dois algoritmos de segmentação.

We present the Mediatrix filamentation method, a novell iterative procedure that decomposes elong... more We present the Mediatrix filamentation method, a novell iterative procedure that decomposes elongated objects in filaments along their main direction over their intensity peak. From this decomposition, the method measures the object's length and thickness. This technique is applied in preliminary tests to arc-shaped objects (simulated gravitational arcs) to recover their curvature center.

Proceedings. XV Brazilian Symposium on Computer Graphics and Image Processing, 2002

This work aims at the implementation of Brazilian vehicle plate localization using image processi... more This work aims at the implementation of Brazilian vehicle plate localization using image processing techniques. In this paper, we show how to find the horizontal position of the plate based on its frequency characteristics in the whole vehicle image. Preliminary results show efficiencies above 96% for images in good illumination conditions.

Proceedings. XV Brazilian Symposium on Computer Graphics and Image Processing, 2002

In image processing, one of the most efficient techniques for image segmentation is entropy-based... more In image processing, one of the most efficient techniques for image segmentation is entropy-based thresholding. In this work a generalized entropy formalism that represents a recent development in statistical mechanics was applied. We propose, for the first time, an image thresholding method using a nonextensive entropy regarding the presence of nonadditive information content in image classes. Preliminary results are shown.

IEEE Latin America Transactions, 2008

Image analysis usually refers to processing of images with the goal of finding objects presented ... more Image analysis usually refers to processing of images with the goal of finding objects presented in the image. Image segmentation is one of the most critical tasks in automatic image analysis. The nonextensive entropy, also known as Tsallis entropy, is a recent development in statistical mechanics and has been considered as a useful measure in describing termostatistical properties of physical systems. In this new formalism a real quantity q was introduced as parameter for physical systems that presents long range interactions, long time memories and fractal-type structures. In image processing, one of the most efficient techniques for image segmentation is entropybased thresholding. This approach uses the Shannon entropy from the information theory considering the gray level image histogram as a probability distribution. In this work, it was applied the Tsallis entropy as a generalized entropy formalism for information theory. For the first time it was proposed an image thresholding method using a nonextensive relative entropy.

Scientometrics, 2009

In this work we have studied the research activity for countries of Europe, Latin America and Afr... more In this work we have studied the research activity for countries of Europe, Latin America and Africa for all sciences between 1945 and November 2008. All the data are captured from the Web of Science database during this period. The analysis of the experimental data shows that, within a nonextensive thermostatistical formalism, the Tsallis q-exponential distribution N (c) satisfactorily describes Institute of Scientific Information citations. The data which are examined in the present survey can be fitted successfully as a first approach by applying a single curve (namely, N (c) ∝ 1/[1 + (q − 1) c/T ] 1 q−1 with q ≃ 4/3 for all the available citations c, T being an "effective temperature". The present analysis ultimately suggests that the phenomenon might essentially be one and the same along the entire range of the citation number. Finally, this manuscript provides a new ranking index, via the "effective temperature" T , for the impact level of the research activity in these countries, taking into account the number of the publications and their citations.

Proceedings 13th Brazilian Symposium on Computer Graphics and Image Processing (Cat. No.PR00878), 2000

This work aims the recognition of characters in plates of private Brazilian vehicles using neural... more This work aims the recognition of characters in plates of private Brazilian vehicles using neural networks. Initially we define a standard for the image as prerequisite for the pre-processing. Afterward, the reduction of the dimensionality is made in order to make possible the use of the network. For this we establish a figure of merit-Q-characterizing, amongst some algorithms of characters splitting, the best one. Q is based on normalized cross correlation of the desired character image and the gotten one. The pre-processing allowed us the improvement of the image, reducing the background noise and increasing the contrast by applying a non-linear function. In the last stage, the neural network recognizes each character of the plate. Preliminary results are shown.