Search

Ecological environments research helps to assess the impacts on forests and managing forests. The usage of novel software and hardware technologies enforces the solution of tasks related to this problem. In addition, the lack of connectivity for large data throughput raises the demand for edge-computing-based solutions towards this goal. Therefore, in this work, we evaluate the opportunity of using a Wearable edge AI concept in a forest environment. For this matter, we propose a new approach to the hardware/software co-design process. We also address the possibility of creating wearable edge AI, where the wireless personal and body area networks are platforms for building applications using edge AI. Finally, we evaluate a case study to test the possibility of performing an edge AI task in a wearable-based environment. Thus, in this work, we evaluate the system to achieve the desired task, the hardware resource and performance, and the network latency associated with each part of the process. Through this work, we validated both the design pattern review and case study. In the case study, the developed algorithms could classify diseased leaves with a circa 90% accuracy with the proposed technique in the field. This results can be reviewed in the laboratory with more modern models that reached up to 96% global accuracy. The system could also perform the desired tasks with a quality factor of 0.95, considering the usage of three devices. Finally, it detected a disease epicenter with an offset of circa 0.5 m in a 6 m × 6 m × 12 m space. These results enforce the usage of the proposed methods in the targeted environment and the proposed changes in the co-design pattern.

In recent years, deep learning methods have outperformed previous state-of-the-art machine learning techniques for several problems, including image classification. Classifying cells in Pap smear images is very challenging, and it is still of paramount importance for cytopathologists. The Pap test is a cervical cancer prevention test that tracks preneoplastic changes in cervical epithelial cells. Carrying out this exam is important in that early detection. It is directly related to a greater chance of curing or reducing the number of deaths caused by the disease. The analysis of Pap smears is exhaustive and repetitive, as it is performed manually by cytopathologists. Therefore, a tool that assists cytopathologists is needed. This work considers 10 deep convolutional neural networks and proposes an ensemble of the three best architectures to classify cervical cancer upon cell nuclei and reduce the professionals’ workload. The dataset used in the experiments is available in the Center for Recognition and Inspection of Cells (CRIC) Searchable Image Database. Considering the metrics of precision, recall, F1-score, accuracy, and sensitivity, the proposed ensemble improves previous methods shown in the literature for two- and three-class classification. We also introduce the six-class classification outcome.

Uma complicação causada pelo diabetes mellitus é o aparecimento de feridas situadas na região dos pés denominadasúlceras do pé diabético. O tratamento tardio pode acarretar o surgimento de infecção ou isquemia da úlcera que, em estado avançado, pode ocasionar a amputação dos membros inferiores. Neste trabalho foi realizado um comparativo do desempenho de diversas arquiteturas de aprendizado profundo pré-treinadas na classificação de imagens deúlceras do pé diabético. Nossa avaliação levou em consideração quatro cenários, três binários Saudável vs Úlceras; Saudável vs Isquemia; Saudável vs Infecção e um multiclasse Saudável vs Úlcera vs Isquemia vs Infecção vs Infecção e Isquemia. Os resultados alcançados demostram que a nossa proposta consegue classificar tais imagens, visto que, nos testes realizados o índice Kappa atingiu valores considerados "Excelentes". Contudo, para o problema multiclasse ainda é necessário aprimorar o uso dessas técnicas.

A natureza subjetiva e interpretativa da análise do exame de Papanicolaou requer grande conhecimento e experiência do profissional responsável. A tecnologia e a construção de ferramentas de auxílio ao citopatologista podem serúteis na busca pela excelência no diagnóstico. Usufruindo de um comportamento usual dos especialistas, de buscar apoio e discussão com outros profissionais, este trabalho propõe o aplicativo CitoFocus como um ambiente colaborativo voltado para a discussão de casos, com foco nas imagens e informações clínicas.

Amidst the current health crisis and social distancing, telemedicine has become an important part of mainstream of healthcare, and building and deploying computational tools to support screening more efficiently is an increasing medical priority. The early identification of cervical cancer precursor lesions by Pap smear test can identify candidates for subsequent treatment. However, one of the main challenges is the accuracy of the conventional method, often subject to high rates of false negative. While machine learning has been highlighted to reduce the limitations of the test, the absence of high-quality curated datasets has prevented strategies development to improve cervical cancer screening. The Center for Recognition and Inspection of Cells (CRIC) platform enables the creation of CRIC Cervix collection, currently with 400 images (1,376 × 1,020 pixels) curated from conventional Pap smears, with manual classification of 11,534 cells. This collection has the potential to advance current efforts in training and testing machine learning algorithms for the automation of tasks as part of the cytopathological analysis in the routine work of laboratories.

Author(s): Silva, RRV; Araujo, FHD; Ushizima, DM; Bianchi, AGC; Carneiro, CM; Medeiros, FNS | Abstract: This paper introduces computational tools for cell classification into normal and abnormal, as well as content-based-image-retrieval (CBIR) for cell recommendation. It also proposes the radial feature descriptors (RFD), which define evenly interspaced segments around the nucleus, and proportional to the convexity of the nuclear boundary. Experiments consider Herlev and CRIC image databases as input to classification via Random Forest and bootstrap; we compare 14 different feature sets by means of False Negative Rate (FNR) and Kappa (κ), obtaining FNR =0.02 and κ=0.89 for Herlev, and FNR =0.14 and κ=0.78 for CRIC. Next, we sort and rank cell images using convolutional neural networks and evaluate performance with the Mean Average Precision (MAP), achieving MAP =0.84 and MAP =0.82 for Herlev and CRIC, respectively. Cell classification show encouraging results regarding RFD, including its sensitivity to intensity variation around the nuclear membrane as it bypasses cytoplasm segmentation.

The leaf damage estimation is a relevant research topic regarding both agricultural and ecological studies. Most approaches lack generality when treating leaves with different shapes. In this work, we propose a procedure to predict the original leaf shape and calculate its defoliation based on a Conditional Generative Adversarial Network (Conditional GAN). We trained and validated the algorithm with a dataset with leaf images from 33 different species. The results indicate the this system's feasibility for the proposed usage.

Specific reading disorders are conditions caused by neurological dysfunctions that affect the linguistic processing of printed text. Many people go untreated due to the lack of specific tools and the high cost of using proprietary software; however, new audio signal processing technologies can help identify genetic pathologies. The methodology developed by medical specialists extracts characteristics from the reading of a text aloud and returns evidence of dyslexia. This work proposes an improvement of the research presented in [25], extracting new features and improvements serving as a tool for dyslexia indication efficiently. The analysis is done in recordings of the reading of pre-defined texts with school-age children. Direct and indirect characteristics of the audio signal are extracted. The direct ones are obtained through the methodology of separation of pauses and syllables. Simultaneously, the indirect characteristics are extracted through the alignment of audio signals, the Hidden Markov Model, and some heuristics of improvement. The indication of the probability of dyslexia is performed using a machine learning algorithm. The tests were compared with the specialist’s classification, obtaining high accuracy on the evidence of dyslexia. The difference between the values of the characteristics collected automatically and manually was below 20% for most features. Finally, the results show a promising research area for audio signal processing concerning the aid to specialists in the decision making related to language pathologies.

Belt Conveyors are the main class of machinery that compose the logistics of a port terminal. The rolling components of the conveyor may fail mainly due to damaged idlers, which may cause a severe industrial breakdown. Nowadays, the equipment protection is done by a set of sensors that indicate an already existing abnormality or by personal inspection, applying practical experience in the search of visual, sound, or temperature signatures of imminent failure. Aiming to upgrade from the current corrective system to the predictive domain, a model for early failure detection on the conveyor’s idlers is proposed. Ultrasound recordings were conducted on idlers that did not present any perceptible abnormalities, labeled as non-defective, and on idlers that displayed typical failure noise, labeled as defectives. The dataset collected was used for the training and testing of Random Forest and Multilayer Perceptron machine learning algorithms. Four types of experiments were devised to test the methodology, two of them using time-domain data, and two of them using frequency domain data, with different statistical attributes. The results achieved in various classification experiments showed that there is a distinctive pattern on the ultrasound spectrum that differs non-defective from defective idlers, as pre-evaluated by traditional methods of human inspection. In the best case, the experiment that used a moving average on the frequency domain data presented an average of 83.68% of correctly classified idlers, obtaining as best result accuracy of 89.47%.

The focus of this work is on the detection of nuclei in synthetic images of cervical cells. Finding nuclei is an important step in building a computational method to help cytopathologists identify cell changes from Pap smears. The method developed in this work combines both the Multi-Start and the Iterated Local Search metaheuristics and uses the features of a region to identify a nucleus. It aims to improve the assertiveness of the screening and reduce the professional workload. The irace package was used to automatically calibrate all parameter values of the method. The proposed approach was compared with other methods in the literature according to recall, precision, and F1 metrics using the ISBI Overlapping Cytology Image Segmentation Challenge database (2014). The results show that the proposed method has the second-best values of F1 and recall, while the accuracy is still high.

In the mining industry, belt conveyors are essential for transporting large quantities of materials efficiently and inexpensively. The rollers are one of the main components of a belt conveyor. Their maintenance and operation are critical for material transportation during the mining activities. Despite that, its inspection has been done in the same way for almost 20 years, with little technological innovation. The inspection of the belt conveyor rollers automatically via object detection and thermal imaging can improve this process. Therefore, we propose a new system capable of running the detection of defective rollers in real time and with better precision and recall metrics than those of previous works. We based the object detection models on the YOLOv2 deep learning architecture. Comparing our results with previous work, we observe reductions on false discovery and false negative rates of 52.46% and 61.03%, respectively. We also evaluated three computing infrastructures, and the GPU-based one was the only that enabled real-time performance.

The Pap test is a preventive approach that requires specialized and labor-intensive examination of cytological preparations to track potentially cancerous cells from the internal and external cervix surface. A cytopathologist must analyze many microscopic fields while screening for abnormal cells. Therefore there is hope that a support decision system could assist with clinical diagnosis, for example, by identifying sub-cellular abnormalities, such as changes in the nuclei features. This work proposes an ensemble method for cervical nuclei detection aiming to reduce the workload of cytopathologists. First, a preprocessing phase divides the original image into superpixels, which are input to feature extraction and selection algorithms. The proposed ensemble method combines three classifiers: Decision Tree (DT), Nearest Centroid (NC), and k-Nearest Neighbors (k-NN), which are evaluated against the ISBI’14 Overlapping Cervical Cytology Image Segmentation Challenge dataset. Experiments show that the proposed method is the state-of-the-art algorithm of the literature for recall (0.999) and F1 values (0.993). It produced a recall very close to the optimum value and also kept high precision (0.988).

In this paper, we introduce and evaluate the systems submitted to the first Overlapping Cervical Cytology Image Segmentation Challenge, held in conjunction with the IEEE International Symposium on Biomedical Imaging 2014. This challenge was organized to encourage the development and benchmarking of techniques capable of segmenting individual cells from overlapping cellular clumps in cervical cytology images, which is a prerequisite for the development of the next generation of computer-aided diagnosis systems for cervical cancer. In particular, these automated systems must detect and accurately segment both the nucleus and cytoplasm of each cell, even when they are clumped together and, hence, partially occluded. However, this is an unsolved problem due to the poor contrast of cytoplasm boundaries, the large variation in size and shape of cells, and the presence of debris and the large degree of cellular overlap. The challenge initially utilized a database of $16$ high-resolution ( $\times$ 40 magnification) images of complex cellular fields of view, in which the isolated real cells were used to construct a database of $945$ cervical cytology images synthesized with a varying number of cells and degree of overlap, in order to provide full access of the segmentation ground truth. These synthetic images were used to provide a reliable and comprehensive framework for quantitative evaluation on this segmentation problem. Results from the submitted methods demonstrate that all the methods are effective in the segmentation of clumps containing at most three cells, with overlap coefficients up to 0.3. This highlights the intrinsic difficulty of this challenge and provides motivation for significant future improvement.

Hyperbilirubinemia is a systemic global problem and an often diagnosed pathology in newborns. In most cases, blue-light phototherapy (410–460 nm) is the only treatment required to prevent the bilirubin neurotoxicity, but its effectiveness is dose-dependent and is highly affected by the skills, knowledge, and attention of nurses and clinicians who implement the treatment. This paper presents a novel light-sensitive colorimetric dosimeter to simulate the effect of blue-light phototherapy on the optical properties of in vitro bilirubin. The dosimeter is based on a multilayered organic structure comprising a light-stable green-light emitter (copper phthalocyanine, C 32 Cl 16 CuN 8 ), and a blue-light-sensitive red emitter (poly{[2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylene]vinylene}, OC 1O C 6 -PPV). The optical and chemical properties of the dosimeter were investigated by fluorescence and FTIR spectroscopies and by color coordinates of CIE (1931) diagram chromatics, while the photochemical processing of bilirubin was evaluated by UV–vis absorption spectroscopy. The optical response of the dosimeter under conventional (10 μW/cm 2 ) and intensive (40 μW/cm 2 ) phototherapy was found to present the same photochemical kinetics as bilirubin. These findings highlight a cutting-edge solution for monitoring bilirubin elimination in neonates as a function of the dosimeter’s color evolution. This novel colorimetric dosimeter provides a standardized method of reporting and measuring phototherapy dose to promote an efficient use of available phototherapy units in health facilities.

This paper describes a variety of experiments that reveal the potential application of impedance spectroscopy (IS) measurements to enhance performance of ultrathin polyaniline (PANI)-based films on flexible substrate to control ammonia gas exposure in the 0–20 ppm range. Further, the device is mechanically robust for making electrical contact without lithography. We have employed the IS technique to proper adjust the optimal operation frequency of an easy-to-process, inexpensive and user-friendly PANI-based resistometric detectors in which drift current and delay problems related to electrode effects and dielectric absorption of PANI are insignificant. These requirements place emphasis upon the material and device configuration, and the range of frequencies from 10 to 100 Hz corresponds to the ideal working region to operate the device to unify sensibility inside 5%, reproducibility, linearity, stability and low response time (

Piloting an unmanned aerial vehicle (UAV), popularly known as drones, is a challenging task that contrasts the pilot's ability and the expertise required for the task. In industrial scenarios, inspectors may not know how to pilot the drone. However, these analysts could potentially assist someone in the process and collaborate on the task. Thus, the purpose of this work is to evaluate a framework that divides the teleoperation of drones into two participants: one participant playing the role of the pilot and another one playing the role of the inspector, over all enabling the teleoperation of drones and performing tasks functioning/working in a cooperative manner. User studies were performed to evaluate the individual opinions of the participants regarding the implemented framework. The results of this work show the framework to be efficient in facilitating the resolution of tasks, which, none the less, could be improved by adding better equipment.

Doenças renais crônicas surgem a partir de patologias agudas ou intermitentes não tratadas adequadamente como a doença de lesão mı́nima (DLM) e a glomeruloesclerose segmentar e focal (GESF). Identificar corre- tamente essas duas doenças é de suma importância, pois seus tratamentos e prognósticos são diferentes. Dessa forma, propomos um método capaz de dife- renciar DLM e GESF através de imagens de exames patológicos. No método proposto, foram extraı́das 10240 caracterı́sticas de três redes neurais convo- lucionais pré-treinadas, foram selecionadas 62 delas através do algoritmo de informação mútua e o Random Forest foi utilizado para a classificação. O método obteve acurácia de 93,33% e Kappa de 85,47%, o que é considerado “Quase Perfeito”.

This paper presents a novel hierarchical nuclei segmentation algorithm for isolated and overlapping cervical cells based on a narrow band level set implementation. Our method applies a new multiscale analysis algorithm to estimate the number of clusters in each image region containing cells, which turns into the input to a narrow band level set algorithm. We assess the nuclei segmentation results on three public cervical cell image databases. Overall, our segmentation method outperformed six state-of-the-art methods concerning the number of correctly segmented nuclei and the Dice coefficient reached values equal to or higher than 0.90. We also carried out classification experiments using features extracted from our segmentation results and the proposed pipeline achieved the highest average accuracy values equal to 0.89 and 0.77 for two-class and three-class problems, respectively. These results demonstrated the suitability of the proposed segmentation algorithm to integrate decision support systems for cervical cell screening.

Prevention of cervical cancer could be performed using Pap smear image analysis. This test screens pre-neoplastic changes in the cervical epithelial cells, accurate screening can reduce deaths caused by the disease. Pap smear test analysis is exhaustive and repetitive work performed visually by a cytopathologist. This article proposes a workload-reducing algorithm for cervical cancer detection based on analysis of cell nuclei features within Pap smear images. We investigate eight traditional machine learning methods to perform a hierarchical classification. We propose a hierarchical classification methodology for computer-aided screening of cell lesions, which can recommend fields of view from the microscopy image based on the nuclei detection of cervical cells. We evaluate the performance of several algorithms against the Herlev and CRIC databases, using a varying number of classes during image classification. Results indicate that the hierarchical classification performed best when using Random Forest as the key classifier, particularly when compared with decision trees, k-NN, and the Ridge methods.

The nuclei and cytoplasm segmentation of cervical cells is a well studied problem. However, the current segmentation algorithms are not robust to clinical practice due to the high computational cost or because they cannot accurately segment cells with high overlapping. In this paper, we propose a method that is capable of segmenting both cytoplasm and nucleus of each individual cell in a clump of overlapping cells. The proposed method consists of three steps: 1) cellular mass segmentation; 2) nucleus segmentation; 3) cytoplasm identification based on an active contour method. We carried out experiments on both synthetic and real cell images. The performance evaluation of the proposed method showed that it was less sensitive to the increase in the number of cells per image and the overlapping ratio against two other existing algorithms. It has also achieved a promising low processing time and, hence, it has the potential to support expert systems for cervical cell recognition.

Frequent and accurate inspections of industrial components and equipment are essential because failures can cause unscheduled downtimes, massive material, and financial losses or even endanger workers. In the mining industry, belt idlers or rollers are examples of such critical components. Although there are many precise laboratory techniques to assess the condition of a roller, companies still have trouble implementing a reliable and scalable procedure to inspect their field assets. This article enumerates and discusses the existing roller inspection techniques and presents a novel approach based on an Unmanned Aerial Vehicle (UAV) integrated with a thermal imaging camera. Our preliminary results indicate that using a signal processing technique, we are able to identify roller failures automatically. We also proposed and implemented a back-end platform that enables field and cloud connectivity with enterprise systems. Finally, we have also cataloged the anomalies detected during the extensive field tests in order to build a structured dataset that will allow for future experimentation.

Due to degradation and low quality in noisy images, such as natural scene images and CAPTCHAs (Completely Automated Public Turing test to tell Computers and Humans Apart) based on text, the character recognition problem continues to be extremely challenging. In this work, we study two convolutional neural network approaches (filter learning and architecture optimization) to improve the feature representations of these images through deep learning. We perform experiments in the widely used Street View House Numbers (SVHN) dataset and a new dataset of CAPTCHAs created by us. The approach to learn filter weights through back-propagation algorithm using data augmentation technique and the strategy of adding few locally-connected layers to the Convolutional Neural Network (CNN) has obtained promising results on the CAPTCHA dataset (97.36% of accuracy for characters and 85.4% for CAPTCHAs) and results very close to the state-of-the-art regarding the SVHN dataset (97.45% of accuracy for digits).

Mobile devices are getting much more relevance during the users' day, in a way that they are paying to increase device security and durability though external cases or insurance plans. However, these approaches are useless if the individuals does not properly take care of their devices. This paper describes an approach to monitor and classifies a surface where a smartphone falls, making possible to categorize this crash into a range of dangerousness. The authors collected empirical data from device falls to make possible the development of an optimal classifier. Our results reached up to 88% recognition rate of surfaces considering a specific features subset, letting us conclude that it is possible to infer user care level through the analysis of how a device is being treated.

The automation process of Pap smear analysis holds the potential to address women's health care in the face of an increasing population and respective collected data. A fundamental step for automating analysis is cell detection from light microscopy images. Such information serves as input to cell classification algorithms and diagnostic recommendation tools. This paper describes an approach to nuclei cell segmentation, which critically impacts the following steps for cell analyses. We developed an algorithm combining clustering and genetic algorithms to detect image regions with high diagnostic value. A major problem when performing the segmentation of images is the cellular overlay. We introduce a new nuclear targeting approach using heuristics associated with a multi-objective genetic algorithm. Our experiments show results using a public 45-image dataset, including comparison to other cell detection approaches. The findings suggest an improvement in the nuclei segmentation and promise to support more sophisticated schemes for data quality control.

A methodology for granulometric image analysis based on modeling the spatial energy of frequencies, using the Wavelet transform is presented in this article. A brief review of the Wavelet Transform tool is presented, and the proposed methodology is detailed. The presented results were obtained using numerically and experimentally simulated images. These results show the relevant correlation between the energy of the Wavelet coefficients and the size distribution of the analyzed objects.

Este trabalho representa continuidade no desenvolvimento de trabalhos na área de neurociência computacional, em particular na área de neuromorfometria e no relacionamento da forma-função. Os objetivos principais são a investigação e a simulação de modelos dinâmicos para o desenvolvimento de células neurais, e a caracterização da sua morfometria em termos de atributos. A tese apresenta um histórico sobre a neurociência, e uma breve revisão sobre a biologia do neurônio e sobre fatores que influenciam na variação na sua forma. Seguimos com a apresentação dos principais modelos computacionais de simulação neural, funcionais e de crescimento neural, com uma descrição mais detalhada de um modelo de crescimento baseado na atuação do cálcio como agente morfogênico e também na polimerização de actinas. Como uma introdução à modelagem neural, discutimos técnicas computacionais de evolução de contornos que podem ser utilizadas na simulação do desenvolvimento neural, propagação de frentes e contornos ativos. Apresentamos também medidas neuromorfométricas tais como a dimensão fractal multiescala, e medidas extraídas a partir do esqueleto da imagem do neurônio, tais como largura, espessura, número de ramos e curvatura das ramificações. Apresentamos os resultados obtidos em diferentes hipóteses de desenvolvimento de células neurais. Foram propostos crescimentos baseados na normal (velocidade na direção normal a curva), convolução, thin plate splines e dinâmica da polimerização da actina. Além disso, foi proposta uma nova abordagem para a evolução da membrana neural baseada em contornos, utilizando a formulação de contornos ativos sob a ação do campo elétrico externo e a curvatura da forma, o que possibilitou a geração de estruturas com características muito semelhantes a do neurônio, inclusive com ramificações. Finalizamos o trabalho apresentando os resultados e conclusões obtidas para os modelos de desenvolvimento. In this thesis we report the investigation and simulation of dynamic models of neural growing, and their characterization using shape features, considering the form function relationship and neuromorphometry. The thesis begins by presenting an overview about neuroscience, neural cell biology and the biological factors that affects the neuron form developments, followed by the presentation of computational neuronal models based on electrophisiological measures and development models of internal structures as actin and microtubules. Special attention is devoted to a neuron growth model based on calcium as a morphogen, whose main characteristic is its electric activity at the membrane. Regarding mathematical models of neural development, two different approaches of contour evolutions are presented, Level Set Methods and Active Contours. Some neuromorphometric measures are implemented and discussed as features for classification and neural evolution, including the multiscale fractal dimension, and dendrite measurements are obtained by using neuron skeletons. In agreement with biological form influences, some hypotheses about development of neuron growth are proposed based on evolution rules, such as: normal evolution (based in normal velocity), convolution, thin plate splines and actin polimerization. A new approach about neuron development is also proposed: a contour based technique that makes use of active contour formulation, Snake Balloon, where the membrane velocity and direction suffers influences of internal and external factors, such as electrical field with diferent geometries, and contour curvature. Both hypotheses are in accordance with the biological factors that influences the neuron form. The simulation produces similar neuron-like structures, even with ramification of certain dendrites

Neste trabalho apresentamos o desenvolvimento e a implementação de diversas técnicas de segmentação de imagens em termos de detecção de bordas, com um destaque especial para a segmentação não-linear. Os métodos considerados foram: o Gradiente, o Laplaciano da Gaussiana, a Regularização linear, e a segmentação não-linear usando o algoritmo Graduated Non Convexity, baseado na minimização de um funcional de energia associado à imagem. O tratamento matemático do funcional foi realizada segundo o paradigma do cálculo variacional. A sua principal vantagem é evidenciada durante o tratamento de bordas e descontinuidades, pois como a segmentação atua de forma não uniforme na imagem, apenas as regiões mais uniformes são suavizadas, preservando as descontinuidades, o que possibilita a conservação mais precisa dos contornos. Nos capítulos destinados a introdução das técnicas computacionais, apresentamos alguns exemplos das segmentações obtidas, possibilitando uma avaliação comparativa e qualitativa dos resultados. Aplicações em micrografias de cristais de KBr e de minerais serviram como um ensaio para a investigação da validação da segmentação através do algoritmo graduated Non Convexity. In this work we describe the development and implementation of several image segmentation techniques, with special attention focused on non linear segmentation. The considered edge detection methods are: Gradient, Laplacian of Gaussian, linear regularization, and the non-linear Graduate Non Convexity segmentation algorithm based on the minimization of the energy functional associated with the image contour. The mathematical treatment was done according to the variational calculus paradigm. The major advantage of such an approach is noted during the treatment of borders and discontinuities, since this method causes the segmentation to act non-uniformelly on the image, in such a way that just the homogeneus regions are smoothed, while preserving discontinuities and enabling more exact localization of the contours. Along the charpters dedicated to introducing the techniques, we present some examples of segmented images, enabling the qualitative and quantitative evaluation of the results. Applications to micrographies of KB4 crystals and minerals in soil provide a possibility to investigate and validate the Graduate Non Convexity segmentation methods.

One potential solution to reduce the concentration of carbon dioxide in the atmosphere is the geologic storage of captured CO2 in underground rock formations, also known as carbon sequestration. There is ongoing research to guarantee that this process is both efficient and safe. We describe tools that provide measurements of media porosity, and permeability estimates, including visualization of pore structures. Existing standard algorithms make limited use of geometric information in calculating permeability of complex microstructures. This quantity is important for the analysis of biomineralization, a subsurface process that can affect physical properties of porous media. This paper introduces geometric and topological descriptors that enhance the estimation of material permeability. Our analysis framework includes the processing of experimental data, segmentation, and feature extraction and making novel use of multiscale topological analysis to quantify maximum flow through porous networks. We illustrate our results using synchrotron-based X-ray computed microtomography of glass beads during biomineralization. We also benchmark the proposed algorithms using simulated data sets modeling jammed packed bead beds of a monodispersive material. © 2012 IEEE.

A careful comparison of three numeric techniques for estimation of the curvature along spatially quantized contours is reported. Two of the considered techniques are based on the Fourier transform (operating over 1D and 2D signals) and Gaussian regularization required to attenuate the spatial quantization noise. While the 1D approach has been reported before and used in a series of applications, the 2D Fourier transform-based method is reported in this article for the first time. The third approach, based on splines, represents a more traditional alternative. Three classes of parametric curves are investigated: analytical, B-splines, and synthesized in the Fourier domain. Four quantization schemes are considered: grid intersect quantization, square box quantization, a table scanner, and a video camera. The performances of the methods are evaluated in terms of their execution speed, curvature error, and sensitivity to the involved parameters. The third approach resulted the fastest, but implied larger errors; the Fourier methods allowed higher accuracy and were robust to parameter configurations. The 2D Fourier method provides the curvature values along the whole image, but exhibits interference in some situations. Such results are important not only for characterizing the relative performance of the considered methods, but also for providing practical guidelines for those interested in applying those techniques to real problems.