Texture can be broadly stated as spatial variation of image intensities. Texture analysis and classification is a well researched area for its importance to many computer vision applications. Consequently, much research has focussed on deriving powerful and efficient texture descriptors. Local binary patterns (LBP) and its variants are simple yet powerful texture descriptors. LBP features describe the texture neighbourhood of a pixel using simple comparison operators, and are often calculated based on varying neighbourhood radii to provide multi-resolution texture descriptions. A comprehensive evaluation of different LBP variants on a common benchmark dataset is missing in the literature. This thesis presents the performance for different LBP variants on texture classification and retrieval tasks. The results show that multi-scale local binary pattern variance (LBPV) gives the best performance over eight benchmarked datasets. Furthermore, improvements to the Dominant LBP (D-LBP) by ranking dominant patterns over complete training set and Compound LBP (CM-LBP) by considering 16 bits binary codes are suggested which are shown to outperform their original counterparts. The main contribution of the thesis is the introduction of multi-dimensional LBP features, which preserve the relationships between different scales by building a multi-dimensional histogram. The results on benchmarked classification and retrieval datasets clearly show that the multi-dimensional LBP (MD-LBP) improves the results compared to conventional multi-scale LBP. The same principle is applied to LBPV (MD-LBPV), again leading to improved performance. The proposed variants result in relatively large feature lengths which is addressed using three different feature length reduction techniques. Principle component analysis (PCA) is shown to give the best performance when the feature length is reduced to match that of conventional multi-scale LBP. The proposed multi-dimensional LBP variants are applied for medical image analysis application. The first application is nailfold capillary (NC) image classification. Performance of MD-LBPV on NC images is highest, whereas for second application, HEp-2 cell classification, performance of MD-LBP is highest. It is observed that the proposed texture descriptors gives improved texture classification accuracy.

UA Open Access Publishing Fund / Introduction: Spinal cord stimulation (SCS) provides significant relief for lumbosacral radiculopathy refractory to both medical and surgical treatment, but historically only offers limited relief for axial low back pain (LBP). We aim to evaluate the response of chronic axial LBP treated with SCS using a surgically implanted epidural paddle lead. Materials and methods: This is a retrospective review of a consecutive series of patients with exclusive LBP or predominant LBP associated with lower extremity (LE) pain evaluated and treated with SCS using an implanted paddle lead within the dorsal thoracic epidural space. Baseline LBP, and if present LE pain, were recorded using the visual analogue scale (VAS) at an initial evaluation. At a follow-up visit (a minimum of 12 months later), LBP and LE pain after a spinal cord stimulator implantation were again recorded using the VAS. Patients were also asked to estimate total LBP pain relief achieved. Results: Patients with either exclusive (n=7) or predominant (n=2) axial LBP were treated with SCS by implantation of a paddle lead at an average spine level of T9. The baseline VAS score for LBP was 7.2; after a follow-up of 20 months, the score decreased to 2.3 (P=0.003). The LE pain VAS score decreased from 7.5 to 0.0 (P=0.103). Patients also reported a subjective 66.4% decrease of their LBP at follow-up. There were no surgical complications. Conclusions: Axial LBP is refractory to many treatments, including SCS. SCS using a surgically implanted paddle electrode provides significant pain relief for chronic axial LPB, and is a safe treatment modality

In a time when cattle herds grow continually larger the need for automatic methods to detect diseases is ever increasing. One possible method to discover diseases is to use thermal images and automatic head and eye detectors. In this thesis an eye detector and a head detector is implemented using the Random Forests classifier. During the implementation the classifier is evaluated using three different descriptors: Histogram of Oriented Gradients, Local Binary Patterns, and a descriptor based on pixel differences. An alternative classifier, the Support Vector Machine, is also evaluated for comparison against Random Forests. The thesis results show that Histogram of Oriented Gradients performs well as a description of cattle heads, while Local Binary Patterns performs well as a description of cattle eyes. The provided descriptor performs almost equally well in both cases. The results also show that Random Forests performs approximately as good as the Support Vector Machine, when the Support Vector Machine is paired with Local Binary Patterns for both heads and eyes. Finally the thesis results indicate that it is easier to detect and locate cattle heads than it is to detect and locate cattle eyes. For eyes, combining a head detector and an eye detector is shown to give a better result than only using an eye detector. In this combination heads are first detected in images, followed by using the eye detector in areas classified as heads.

Ingeniero Civil Eléctrico / Un sistema de identificación facial consiste en un método que toma una imagen del rosto de un individuo desconocido y un banco de imágenes de personas conocidas, con el objetivo de identificar a este individuo y decidir si está presente o no en el banco de imágenes conocidas. Este Trabajo de Titulo se propone como objetivo estudiar el uso de escalas de alta resolución y combinaciones de las mismas en el problema de reconocimiento frontal de rostros. Para esto se utiliza la base de datos XM2VTS que contiene imágenes de rostros de individuos a una resolución de 720x576 píxeles. Se propone generar imágenes a escalas del 50 % y 25% de escala de resolución mediante el método de Pirámides de Gauss, obteniendo tres distintas resoluciones. Luego son procesadas mediante LBP y Filtros Gabor para realizar combinaciones de la información extraída por estos métodos en distintas escalas de resolución. Finalmente se utiliza este vector de información en un clasificador K-NN y se logra el reconocimiento de cada individuo. Se realizan pruebas de reconocimiento para realizar el ajuste de parámetros de los métodos de extracción de características en distintas escalas de resolución y pruebas para detectar las combinaciones de información de estas últimas que entreguen los mejores resultados. Por separado los métodos LBP y Filtros Gabor obtuvieron como máximos desempeños un 97,96% y un 94,12% de reconocimiento exitoso respectivamente. Sin embargo, al fusionar la información obtenida a través de estas técnicas de extracción de características, se obtiene un método de reconocimiento frontal de rostros que logra un 98,9% de reconocimiento exitoso, con un tiempo de procesamiento total de 21 minutos y 1,26 segundos por cada individuo como resultado final. Comparando con trabajos similares realizados con las mismas muestras se posiciona a la par de estos, demostrando ser un método competente y eficiente, además de entregar pautas para seguir el estudio del problema del reconocimiento facial.

Title: Pelvis position in 3D space and its infuence of stages of pregnancy Summary: The aim of this master thesis is to evaluate changes of the pelvis position of women during pregnancy and if these changes influence the LBP during gravidity. The research is based on the summary of theoretical findings about the issue of changes during pregnancy which could be responsible for the LBP. The main topics of this part of thesis are anatomy and kinesiology of pelvis, functional vertebral disorders, LBP in gravidity, physiological and biomechanical changes during pregnancy, definition of 3D space with connection to human body and introduction of the 3D kinematic analysis. 10 (9) pregnant women in the age of 26-35 years was evaluated in 3rd , 5th and 8th month of gravidity. The pelvis position in 3D space was objectified with the Qualisys technology and the LBP and the PGP with subjective evaluation of the VAS. LBP was found in 7 of 10 and PGP in 3 of 10 evaluated pregnant women. Our findings show the tendency to the change of the pelvis position in the sagittal plane in the direction of the forward tilt. In our group of evaluated woman was found the connection between the absolute change of the pelvis position in the sagittal plane and the LBP and the PGP during pregnancy, but to define clear and...

O greening ou Huanglongbing (HLB) é uma das mais graves doenças dos citros presentes nos pomares do Brasil. Causada pela bactéria Candidatus Liberibacter spp, é transmitida pelo inseto psilídeo Diaphorina citri, que ao se alimentar de uma planta doente transmite a doença às demais plantas. O greening apresenta como sintoma, manchas amareladas nas folhas, muitas vezes confundidas com deficiências nutricionais. A erradicação da planta e o controle do inseto transmissor são as únicas formas de prevenção para evitar a sua propagação. Este trabalho teve por objetivo avaliar uma metodologia baseada em segmentação por cor e outra baseada em análise de textura para avaliação de folhas de citros sintomáticas, identificando se estão contaminadas com o greening ou outras doenças e deficiências nutricionais. Foram fornecidas pelo grupo FISHER, 324 amostras de folhas cítricas, contendo folhas com doenças (greening, CVC e rubelose) e deficiências nutricionais (manganês, magnésio e zinco). As folhas foram digitalizadas por um scanner de mesa, com duas resoluções, utilizando somente a parte frontal da folha. Foram montados três bancos de imagens. Os resultados gerados com a metodologia baseada em segmentação por cor utilizando RNA PMC, mostraram que essa metodologia não é eficiente. Na metodologia baseada na análise por textura foram avaliados os descritores LBP, LFP e os de Haralick. Para estes descritores foram extraídas amostras por folha e por quadrantes das folhas nos canais de cores vermelho e verde e amostras em níveis de cinza. Os resultados gerados pelos descritores foram classificados pela distância &#9672 e pelos algoritmos IBK e RNA PMC do toolbox Weka. Os melhores resultados foram para os descritores LBP e LFP-s para distância &#9672, com valores de sensibilidade acima de 97% e 93%, respectivamente, e para o LBP com o algoritmo IBK, com valores de sensibilidade acima de 98,5%. Os resultados obtidos evidenciam que o descritor LBP é o mais eficiente seguido pelo LFP-s na diferenciação do greening das outras pragas. / The greening or Huanglongbing (HLB) is one of the most serious diseases of citrus orchards present in Brazil. HLB is caused by the bacterium Candidatus Liberibacter spp, it is transmitted by the psyllid insect (Diaphorina citri) that, when feeding on a diseased plant, it transmits the disease to other plants. One of the symptoms of the greening are yellowish spots on the leaves, often confused with nutritional deficiencies. The eradication of plants and control of insect are the only forms of prevention. This work aims to evaluate two methodologies: one based on color segmentation and the other based on texture analysis for assessment of symptomatic citrus leaves, identifying whether they are infected with greening and other diseases and nutritional deficiencies. A number of 324 samples of citrus leaves were provided by FISHER group, infected with diseases (greening, CVC, rubelose) and nutritional deficiencies ( manganese, magnesium, zinc) . The leaves were acquired by a flatbed scanner with two different resolutions, using only the front side of the leaf. Three datasets of images were constructed. The results generated using the methodology based on color segmentation with ANN MLP, showed that this methodology is not efficient. In the methodology based on texture analysis it was evaluated the LBP, LFP and the Haralick descriptors. For these descriptors it was extracted samples from the leaves and quadrants of leaves, in red and green color channels and grayscale. The results generated by the descriptors were classified by &#9672 distance and the algorithms IBK and ANN MLP from the toolbox Weka. The best results were for LBP descriptor and LFP-s for &#9672 distance with values of sensitivity above 97% and 93%, respectively, and the LBP with IBK algorithm, with values of sensitivity above 98.5%. The results showed that the LBP descriptor is the most efficient followed by LFP-s in the differentiation of the greening from other pests.

O citoesqueleto é a estrutura celular mais importante em células eucariotas e é responsável por manter a forma da célula e as junções celulares, auxiliando nos movimentos celulares. Esta é composta de filamentos de Actina, Microtúbulos e filamentos intermediários. Recentemente, a análise de duas dessas estruturas tornaram-se importantes, pois é possível obter micrografias usando microscópios de alta resolução, que contém microscopia de fluorescência, em combinação com métodos complexos de aplicação de substâncias de contraste para rotulagem e posterior análises visuais. A combinação dessas técnicas, entretanto, limita-se a ser descritiva e subjetiva. Neste trabalho, são avaliadas cinco técnicas de análise de imagens, as quais são: Bag of Visual Words (BoVW), Local Binary Local (LBP), Textons baseados em Discrete Fourier Transform (TDFT), Textons baseados em Gabor Filter Banks (TGFB) e Textons baseados em Complex Networks (TCN) sobre o conjunto de dados 2D Hela e FDIG Olympus. Experimentos extensivos foram conduzidos em ambos os conjuntos de dados, e seus resultados podem servir de base para futuras pesquisas como análises do citoesqueleto em imagens de microscopia fluorescente. Neste trabalho, é apresentada uma comparação quantitativa e qualitativa dos métodos acima mencionados para entender o comportamento desses métodos e propriedades dos microfilamentos de actina (MA) e Microtúbulos (MT) em ambos os conjuntos de dados. Os resultados obtidos evidenciam que é possível classificar o conjunto de dados da FDIG Olympus com uma precisão de até 90:07% e 98:94% para 2D Hela, além de obter 86:05% e 96:84%, respectivamente, de precisão, usando teoria de redes complexas. / The cytoskeleton is the most important cellular structure in eukaryotic cells and is responsible for maintaining the shape of the cell and cellular junctions, aiding in cell movements. This is composed of filaments of Actin, Microtubules and intermediate filaments. Recently, the analysis of two of these structures has become important because it is possible to obtain micrographs using microscopes of high resolution and fluorescence technology, in combination with complex methods of application of substances of contrast for labeling and later visual analysis. The use of these techniques, however, is limited to being descriptive and subjective. In this work, we evaluate some of the most popular image analysis techniques such as Bag of Visual Words (BoVW), Local Binary Pattern (LBP), Textons based on Discrete Fourier Transform(TDFT) , Gabor Filter banks (TGFB), and approaches based on Complex Networks theory (TCN) over the famous dataset 2D Hela and FDIG Olympus. Extensive experiments were conducted on both datasets in which their results can serve as a baseline for future research with cytoskeleton classification in microscopy fluorescence images. In this work, we present the quantitative and qualitative comparison of above mentioned methods for better understand the behavior of these methods and the properties of Actin microfilaments (MA) and Microtubules (MT) on both datasets. The results showed that it is possible to classify the FDIG Olympus data set with accuracy of up to 90:07% and 98:94% for 2D Hela, in addition to reaching 86:05% and 96:84% respectively, using complex network theory.

O greening ou Huanglongbing (HLB) é uma das mais graves doenças dos citros presentes nos pomares do Brasil. Causada pela bactéria Candidatus Liberibacter spp, é transmitida pelo inseto psilídeo Diaphorina citri, que ao se alimentar de uma planta doente transmite a doença às demais plantas. O greening apresenta como sintoma, manchas amareladas nas folhas, muitas vezes confundidas com deficiências nutricionais. A erradicação da planta e o controle do inseto transmissor são as únicas formas de prevenção para evitar a sua propagação. Este trabalho teve por objetivo avaliar uma metodologia baseada em segmentação por cor e outra baseada em análise de textura para avaliação de folhas de citros sintomáticas, identificando se estão contaminadas com o greening ou outras doenças e deficiências nutricionais. Foram fornecidas pelo grupo FISHER, 324 amostras de folhas cítricas, contendo folhas com doenças (greening, CVC e rubelose) e deficiências nutricionais (manganês, magnésio e zinco). As folhas foram digitalizadas por um scanner de mesa, com duas resoluções, utilizando somente a parte frontal da folha. Foram montados três bancos de imagens. Os resultados gerados com a metodologia baseada em segmentação por cor utilizando RNA PMC, mostraram que essa metodologia não é eficiente. Na metodologia baseada na análise por textura foram avaliados os descritores LBP, LFP e os de Haralick. Para estes descritores foram extraídas amostras por folha e por quadrantes das folhas nos canais de cores vermelho e verde e amostras em níveis de cinza. Os resultados gerados pelos descritores foram classificados pela distância &#9672 e pelos algoritmos IBK e RNA PMC do toolbox Weka. Os melhores resultados foram para os descritores LBP e LFP-s para distância &#9672, com valores de sensibilidade acima de 97% e 93%, respectivamente, e para o LBP com o algoritmo IBK, com valores de sensibilidade acima de 98,5%. Os resultados obtidos evidenciam que o descritor LBP é o mais eficiente seguido pelo LFP-s na diferenciação do greening das outras pragas. / The greening or Huanglongbing (HLB) is one of the most serious diseases of citrus orchards present in Brazil. HLB is caused by the bacterium Candidatus Liberibacter spp, it is transmitted by the psyllid insect (Diaphorina citri) that, when feeding on a diseased plant, it transmits the disease to other plants. One of the symptoms of the greening are yellowish spots on the leaves, often confused with nutritional deficiencies. The eradication of plants and control of insect are the only forms of prevention. This work aims to evaluate two methodologies: one based on color segmentation and the other based on texture analysis for assessment of symptomatic citrus leaves, identifying whether they are infected with greening and other diseases and nutritional deficiencies. A number of 324 samples of citrus leaves were provided by FISHER group, infected with diseases (greening, CVC, rubelose) and nutritional deficiencies ( manganese, magnesium, zinc) . The leaves were acquired by a flatbed scanner with two different resolutions, using only the front side of the leaf. Three datasets of images were constructed. The results generated using the methodology based on color segmentation with ANN MLP, showed that this methodology is not efficient. In the methodology based on texture analysis it was evaluated the LBP, LFP and the Haralick descriptors. For these descriptors it was extracted samples from the leaves and quadrants of leaves, in red and green color channels and grayscale. The results generated by the descriptors were classified by &#9672 distance and the algorithms IBK and ANN MLP from the toolbox Weka. The best results were for LBP descriptor and LFP-s for &#9672 distance with values of sensitivity above 97% and 93%, respectively, and the LBP with IBK algorithm, with values of sensitivity above 98.5%. The results showed that the LBP descriptor is the most efficient followed by LFP-s in the differentiation of the greening from other pests.

O citoesqueleto é a estrutura celular mais importante em células eucariotas e é responsável por manter a forma da célula e as junções celulares, auxiliando nos movimentos celulares. Esta é composta de filamentos de Actina, Microtúbulos e filamentos intermediários. Recentemente, a análise de duas dessas estruturas tornaram-se importantes, pois é possível obter micrografias usando microscópios de alta resolução, que contém microscopia de fluorescência, em combinação com métodos complexos de aplicação de substâncias de contraste para rotulagem e posterior análises visuais. A combinação dessas técnicas, entretanto, limita-se a ser descritiva e subjetiva. Neste trabalho, são avaliadas cinco técnicas de análise de imagens, as quais são: Bag of Visual Words (BoVW), Local Binary Local (LBP), Textons baseados em Discrete Fourier Transform (TDFT), Textons baseados em Gabor Filter Banks (TGFB) e Textons baseados em Complex Networks (TCN) sobre o conjunto de dados 2D Hela e FDIG Olympus. Experimentos extensivos foram conduzidos em ambos os conjuntos de dados, e seus resultados podem servir de base para futuras pesquisas como análises do citoesqueleto em imagens de microscopia fluorescente. Neste trabalho, é apresentada uma comparação quantitativa e qualitativa dos métodos acima mencionados para entender o comportamento desses métodos e propriedades dos microfilamentos de actina (MA) e Microtúbulos (MT) em ambos os conjuntos de dados. Os resultados obtidos evidenciam que é possível classificar o conjunto de dados da FDIG Olympus com uma precisão de até 90:07% e 98:94% para 2D Hela, além de obter 86:05% e 96:84%, respectivamente, de precisão, usando teoria de redes complexas. / The cytoskeleton is the most important cellular structure in eukaryotic cells and is responsible for maintaining the shape of the cell and cellular junctions, aiding in cell movements. This is composed of filaments of Actin, Microtubules and intermediate filaments. Recently, the analysis of two of these structures has become important because it is possible to obtain micrographs using microscopes of high resolution and fluorescence technology, in combination with complex methods of application of substances of contrast for labeling and later visual analysis. The use of these techniques, however, is limited to being descriptive and subjective. In this work, we evaluate some of the most popular image analysis techniques such as Bag of Visual Words (BoVW), Local Binary Pattern (LBP), Textons based on Discrete Fourier Transform(TDFT) , Gabor Filter banks (TGFB), and approaches based on Complex Networks theory (TCN) over the famous dataset 2D Hela and FDIG Olympus. Extensive experiments were conducted on both datasets in which their results can serve as a baseline for future research with cytoskeleton classification in microscopy fluorescence images. In this work, we present the quantitative and qualitative comparison of above mentioned methods for better understand the behavior of these methods and the properties of Actin microfilaments (MA) and Microtubules (MT) on both datasets. The results showed that it is possible to classify the FDIG Olympus data set with accuracy of up to 90:07% and 98:94% for 2D Hela, in addition to reaching 86:05% and 96:84% respectively, using complex network theory.

Både ospecifik akut- och kronisk ländryggssmärta (LBP) är stora problem världen över, både för den generella populationen och elitidrottare. Elitidrottare ställs för höga fysiska påfrestningar som kan påverka LBP. I studien undersöks prevalens av LBP hos elithandbollsspelare samt eventuella samband mellan enskilda variabler såsom exempelvis träningsmängd och spelpositioner och LBP hos elithandbollsspelare. Utöver detta undersöks även spelarnas bedömning av huruvida LBP påverkar deltagandet i idrotten. En webbaserad enkät skickades ut till elitklubbar i Handbollsligan, SHE och Allsvenskan. I studien deltog 91 individer, 46 män och 45 kvinnor. Frågorna i enkäten behandlade prevalens av LBP samt variabler som kan ha ett samband med LBP. Totalt hade 77% besvärats av LBP och 40% av dessa lider av kronisk LBP. En högre medianålder var signifikant associerat med LBP (p=0,044). Inga signifikanta skillnader syntes i fördelningstester mellan prevalent LBP och spelstil (kvinnor p=0,6 män p=0,4 totalt p=0,8), LBP och anfallsposition (kvinnor p=0,8 män p=0,3 totalt p=0,5), LBP och försvarsposition (kvinnor p=0,2 män p=0,8 totalt p=0,7). Logistisk regression påvisade inget signifikant samband mellan träningsmängd och LBP (p=0,1, odds-ratio 0,574, CI 0,296 – 1,11). Justering för ålder påverkade inte oddsen för LBP nämnvärt (p=0,1, odds-ratio 0,581, CI 0,294 – 1,14). Det syntes en stor diskrepans i huruvida spelarna missade träning och matcher jämfört med hur LBP påverkade deltagandet överlag. Inga signifikanta samband mellan spelstil, spelposition eller träningsmängd och LBP påvisades i studien. Det visas dock att LBP är vanligt förekommande inom svensk elithandboll men att deltagarantalet var för litet för att dra säkra slutsatser kring detta. / Both unspecific acute and chronic low-back pain (LBP) are global problems in the general public aswell as elite athletes. Elite-athletes are exposed to high physical demands which could affect LBP. This study investigated the prevalence of LBP in elite-handballers, associations between LBP and several variables such as handball-sessions/week and playing-position. Secondly, the study investigated if LBP impedes on players participation within sport. A web-based survey was sent to elite-clubs in Sweden’s highest divisions; Handbollsligan, SHE, Allsvenskan. A total of 91 players participated, 46 men and 45 women. The survey-questions investigated the prevalence of LBP and variables which could be associated with LBP. A total of 77% stated issues with LBP, 40% of these suffers from chronic LBP. A higher median age were associated with LBP (p=044). No statistical significance was found in Chi-2 tests between prevalent LBP and style of play (women p=0,6 men p=0,6 total p=0,8), LBP and offensive-position (women p=0,8 Men p=0,3 total p=0,5), LBP and defensive-position (women p=0,2 men p=0,8 total p=0,7). No significant association was found between handball-sessions/week and LBP (p=0,1, odds-ratio 0,574, CI 0,296 – 1,11). Adjustment for age did not affect the outcome mentionably (p=0,1, odds-ratio 0,581, CI 0,294 – 1,14). Many players experienced that LBP affected their participation overall but few missed matches and training because of LBP. No significant association between style of play, playing-position or handball-sessions and LBP was found. LBP is common in swedish elite-handballers, but the number of participants were too low to state this for certain.