Made available in DSpace on 2014-06-11T19:22:33Z (GMT). No. of bitstreams: 0 Previous issue date: 2008-06-25Bitstream added on 2014-06-13T18:49:33Z : No. of bitstreams: 1 silva_jf_me_ilha.pdf: 1508098 bytes, checksum: a71768d52f587fb0eddda6d0c43e13e9 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Neste trabalho é apresentado um sistema interativo para o processamento e armazenamento de imagens no domínio wavelet e uma interface para processamento de imagens digitais. A proposta apresentada baseia-se na transformada wavelet e em métodos de limiar. O sistema de armazenamento tem como objetivo a otimização do espaço computacional, tanto para armazenamento como para transmissão de imagens. Sendo para isso necessário a aplicação da Transformada Wavelet nas respectivas imagens e dos diversos métodos de limiar a serem escolhidos. Estas aplicações permitem extrair informações relevantes para o armazenamento de uma imagem com um menor custo computacional e com uma margem de erro muito pequena quando se compara as imagens, original e processada, ou seja, não há perda de qualidade ao aplicar o sistema de codificação apresentado. Os resultados obtidos a partir das informações extraídas das imagens são apresentados numa interface gráfica. É através da interface gráfica que o usuário usa os arquivos para visualizar e analisar os resultados dos programas diretamente na tela do computador sem a preocupação de lidar com os códigos fontes. A interface gráfica, os programas de processamento de imagens via Transformada Wavelet e os métodos de limiar foram desenvolvidos no ambiente do MATLAB, possibilitando uma troca direta de informações entre eles e o usuário. / In this work an interactive system processing and storage of compressed images in the wavelet domain, and an interface for digital image processing are presented. The storage method proposed is based on the wavelet transform and a threshold method. It has the objective of optimizing the computational space, for storage and for transmission of images. It is necessary first a pre-processing routine and after, the characterization through the application of the wavelet transform to the respective images and the several threshold methods to be chosen. These applications allow the extraction of relevant information for an image compression or storage image with a smaller computational cost and with a very small of error when the original and processed images are compared, in other words, there is no quality loss when applying the codification system presented. The results obtained from the extracted information of the images are presented in a graphic interface for image processing. The graphic interface, the programs for image processing using the wavelet transform and the thresholding methods were developed in a MATLAB environment, making possible a direct change of information between them and the user, without the concern of working with source codes. Through this graphic interface the user can use files which permits him visualize and analyze the results of the programs in the computer screen.

For demanding applications, for example image or video processing, there may be computations that aren’t very suitable for digital signal processors. While a DSP processor is appropriate for some tasks, the instruction set could be extended in order to achieve higher performance for the tasks that such a processor normally isn’t actually design for. The platform used in this project is flexible in the sense that new hardware can be designed to speed up certain computations. This thesis analyzes the computational complex parts of JPEG2000. In order to achieve sufficient performance for JPEG2000, there may be a need for hardware acceleration. First, a JPEG2000 decoder was implemented for a DSP processor in assembler. When the firmware had been written, the cycle consumption of the parts was measured and estimated. From this analysis, the bottlenecks of the system were identified. Furthermore, new processor instructions are proposed that could be implemented for this system. Finally the performance improvements are estimated.

AssociaÃÃo Teresinense de Ensino / Este trabalho tem por objetivo implementar um mÃtodo de localizaÃÃo de faltas em linhas de transmissÃo utilizando sinais de ondas viajantes e transformada wavelet. Para avaliar o desempenho do sistema desenvolvido, consideraÃÃes sobre as diversas famÃlias de wavelet e a influÃncia do ruÃdo branco sÃo investigadas. Para investigar o problema da localizaÃÃo de faltas com sinais de alta frequÃncia utilizando a teoria das ondas viajantes, uma linha de 500kV do Sistema EletrobrÃs - CHESF que interliga as subestaÃÃes de Teresina II e Sobral III à simulada atravÃs do Software ATP (Alternalive Transient Program). Na simulaÃÃo, diferentes tipos de contingÃncias sÃo utilizados com sinais amostrados na frequÃncia de 400kHz. Os sinais provenientes das simulaÃÃes do ATP sÃo processados atravÃs do algoritmo de localizaÃÃo de faltas fundamentado na determinaÃÃo do tempo de viagem das ondas de tensÃo do ponto de falta aos terminais da linha. Para a determinaÃÃo dos instantes de viagem das ondas utiliza-se a transformada wavelet com a anÃlise multiresoluÃÃo wavelet (AMR) em um nÃvel de decomposiÃÃo. Uma vez determinado os intervalos de reflexÃo das ondas, a distÃncia da falta à estimada em funÃÃo destes intervalos e da velocidade de propagaÃÃo da onda na linha. Jà o ruÃdo branco à adicionado aos sinais de tensÃo com intuito de aproximar o sinal proveniente de simulaÃÃes computacionais dos sinais encontrados em oscilografias reais. Os resultados alcanÃados demonstram que o algoritmo de localizaÃÃo apresentou Erro MÃdio Relativo Total (EMRT) aceitÃvel para RelaÃÃo Sinal RuÃdo (Signal-to-Noise Ratio â SNR) a partir de 60dB. Destaque especial para a wavelet rbio 3.5 por apresentar os melhores resultados em todos os tipos de faltas considerados com SNR a partir de 60dB com EMRT de 0,15%, representando um erro absoluto mÃdio de 500m. Assim, os resultados alcanÃados demonstram uma otimizaÃÃo de performance em razÃo da escolha da wavelet mais adequada ao algoritmo e norteiam para uma aplicaÃÃo prÃtica do localizador de faltas. / This work aims to implement a method of fault location in transmission lines making use of traveling waves and wavelet transform. To evaluate the performance of the developed system, considerations on the several families of wavelet and the influence of the white noise are investigated. To observe the problem of fault location with high frequency signals making use of the theory of traveling waves, a 500kV line from the Eletrobras System â CHESF that links Teresina II and Sobral III sub-stations is simulated through ATP (Alternative Transient Program) software. In that simulation, different kinds of contingent are used with sample signs in 400 kHz. The signals that came from the ATP simulations were processed through the algorithm fault location based on the traveling time definition of the tension waves from the fault point on the line terminals. To determine the moment of the wave travels, it was used the wavelet transform with the wavelet multiresolution analysis (WMA) in a decomposition level. Since the Wave Reflection Intervals were defined, the fault distance was estimated taking into account those intervals and speed of propagation of the wave on the line. The white noise is added to the tension signs aiming to bring forward the sign from the computer simulations form the signs found in actual oscilographies. The results have shown that that the localization algorithm had Total Relative Average Error (EMRt) acceptable for the Signal-to-Noise Ratio (SNR) from 60dB. Special highlight must be made to the wavelet rbio 3.5 because it showed the best results in all kinds of faults considered with SNR from 60dB with 0.15% EMRt, which represents a 500m absolute average error. Thus, the results show a performance optimization regarding the wavelet that suits best to the algorithm and guides practical application of the fault locator.

Tendo em vista o papel que as chuvas exercem sobre águas continentais e a consequente disponibilidade hídrica para abastecimento, geração de energia, produção de alimentos e a dinâmica climatológica regional, o estudo de periodicidades nas variações em séries temporais pluviométricas têm impactos relevantes para o planejamento dos recursos hídricos. A precipitação, enquanto elemento do sistema climático, está em função de fatores internos e forçantes extraterrestres. Dentre estas, a força da gravidade da lua tem papel significativo como motor das marés oceânicas e sua influência na distribuição da massa e calor em escala global. Com o objetivo de identificar longos períodos de variação hidroclimática, esse trabalho investigou ciclos em escala decenal na variabilidade de seis séries longas de precipitação na região hidrográfica do Paraná no Sudeste Brasileiro, e sua potencial relação com o ciclo nodal de 18,6 anos. Para alcançar esse objetivo, além das análises estatísticas clássicas de dados, foi utilizado o filtro HP com o intuito de separar componentes de cíclicas e de longo prazo. Foi utilizada a metodologia de análise espectral de dados, como Fourier e a transformada em wavelet, que determina o poder das componentes dentro de uma série temporal. As análises estatísticas e o filtro HP foram importantes na visualização e detecção de períodos persistentes de desvios em relação aos valores médios. A ferramenta de transformada em wavelet apontou para robusta presença de periodicidades decenais em todas as séries, sendo as mais significativas 9 anos e 19 anos para todas as séries, com significância superior a 95%. Os resultados sugerem uma aparente anticorrelação, ou seja, o mínimo do ciclo nodal coincide com um incremento nos totais pluviométricos na região. Assim, essas informações da influência podem aprimorar a previsibilidade de condições hidrológicas interdecenais médias e aquelas variáveis socioeconômicas que são sensíveis à precipitação. / Considering the role of rainfall on continental waters and resulting water availability for water supply, energy generation, food production and regional climatological dynamics, the study of periodicities in variations in rainfall time series has a relevant impact on resource planning Water resources. Precipitation, as an element of the climate system, is a function of internal factors and extraterrestrial forcing. Among these, the force of gravity of the moon plays a significant role as the motor of the ocean tides and its influence on the distribution of mass and energy on a global scale. In order to identify long periods of hydroclimatic variation, this work investigated decadal scale cycles in the variability of six long series of precipitation in the Paraná hydrographic region in southeastern Brazil, and its potential relation with the nodal cycle of 18.6 years. To achieve this goal, in addition to the classic statistical data analysis, the HP filter was used in order to separate components from cyclical/random and long term. We use the spectral data analysis methodology, such as Fourier and the wavelet transform, which determines the power of the components within a time series. Statistical analyzes and the HP filter were important in visualizing and detecting persistent periods of deviations from mean values. The wavelet transform analysis pointed to robust presence of decadal periodicities in all series, being the most significant 9 years and 19 years for all series, with significance higher than 95%. The results suggest an apparent anticorrelation, that is, the minimum of the nodal cycle coincides with an increase in the total rainfall in the region. Thus, this influence information can enhance the predictability of medium interdecadal hydrological conditions and those socioeconomic variables that are sensitive to precipitation.

O presente trabalho visa a utilização da transformada Wavelet no monitoramento do sistema elétrico no que diz respeito a problemas de qualidade da energia com o intuito de detectar, localizar e classificar os mesmos. A transformada Wavelet tem surgido na literatura como uma nova ferramenta para análise de sinais, utilizando funções chamadas Wavelet mãe para mapear sinais em seu domínio, fornecendo informações simultâneas nos domínios tempo e freqüência. A transformada Wavelet é realizada através de filtros decompondo-se um dado sinal em análise multiresolução. Por esta, obtém-se a detecção e a localização de distúrbios relacionados com a qualidade da energia decompondo-se o sinal em dois outros que representam uma versão de detalhes (correspondente as altas freqüências do sinal) e uma versão de aproximação (correspondente as baixas freqüências do sinal). A versão de aproximação é novamente decomposta obtendo-se novos sinais de detalhes e aproximações e assim sucessivamente. Sendo assim, os distúrbios podem ser detectados e localizados no tempo em função do seu conteúdo de freqüência. Estas informações fornecem também características únicas pertinentes a cada distúrbio, permitindo classificá-los. Desta forma, propõe-se neste trabalho o desenvolvimento de um algoritmo classificador automático de distúrbios relacionados com a qualidade da energia baseado unicamente nas decomposições obtidas da análise multiresolução. / The aim of the present dissertation is to apply the Wavelet transform to analyze power quality problems, detecting, localizing and classifying them. The topic Wavelet transform, has appeared in the literature as a new tool for the analysis of signals, using functions called mother Wavelet to map signals in its domain, supplying information in the time and frequency domain, simultaneously. Wavelet transform is accomplished through filters decomposing a provided signal in multiresolution analysis. The detection and localization of disturbances are obtained by decomposing a signal into two other signals that represent, a detailed version (high frequency signals) and a smoothed version (low frequency signals). The smoothed version is decomposed again, and new detailed and smoothed signals are obtained. This process is repeated as many times as necessary and the disturbances can be detected and localized in the time as a function of its level frequency. This information also supplies characteristics to each disturbance, allowing classifying them. Thus, this research presents a way to develop an automatic classifying algorithm of power quality disturbances, based only on multiresolution analysis.

Medical image is a core step of medical diagnosis and has been diffusely applied in modern medical domain. The technology of modern medical image is more and more mature which could present images in different modes and features. Medical image fusion is the technology that could compound two mutual images into one according to certain rules to achieve clear visual effect. By observing medical fusion image, doctor could easily confirm the position of illness. According to the mutual features of CT medical image and MRI medical image, based on the technology of wavelet transform, the paper presents twp effective and applied medical image fusion methods. The first method is based on the features of certain area. The principle is to construct weighted factor and matching degree with certain related parameters to compound the area of high frequency which presents the detailed information. To the area of low frequency, principle of maximum absolute value is selected. Finally we get the fusion image by wavelet reconfiguration. By estimate of subjectivity and objectivity, the method is applied that could export excellent visual effect and good parameters. The other method is based on lifting wavelet. It decomposes the original image to area of low frequency and high frequency, and then transforms them with different fusion rules. To area of low frequency, weighted fusion is applied and to area of high frequency, rule of maximum standard deviation is chosen. Finally we get fusion image from wavelet reconstruction. By the estimate of subjectivity and objectivity, the method is an applied and excellent way that keeps the detailed information effectively and presents clear profile. At the same time, the executed time is shorter than others.

We propose a recursive formula for moments of scaling function and sum rule. It is shown that some quadrature formulae has a higher degree of accuracy under proposed moment condition. On this basis we obtain higher accuracy formula for wavelet expansion coefficients which are needed to start the fast wavelet transform and estimate convergence rate of wavelet approximation and sampling of smooth functions. We also present a direct algorithm for solving refinement equation.

Digitization of medical data has become a very important part of the modern healthcare system. Data can be transmitted easily at any time to anywhere in the world using Internet to get the best diagnosis possible for a patient. This digitized medical data must be protected at all times to preserve the doctor-patient confidentiality. Watermarking can be used as an effective tool to achieve this. In this research project, image watermarking is performed both in the spatial domain and the frequency domain to embed a shared image with medical image data and the patient data which would include the patient identification number. For the proposed system, Structural Similarity (SSIM) is used as an index to measure the quality of the watermarking process instead of Peak Signal to Noise Ratio (PSNR) since SSIM takes into account the visual perception of the images compared to PSNR which uses the intensity levels to measure the quality of the watermarking process. The system response under ideal conditions as well as under the influence of noise were measured and the results were analyzed.

Kyoto University (京都大学) / 0048 / 新制・課程博士 / 博士(工学) / 甲第17106号 / 工博第3617号 / 新制||工||1549(附属図書館) / 29836 / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 杉浦 邦征, 教授 白土 博通, 教授 河野 広隆 / 学位規則第4条第1項該当

This work applies sparse representations and nonlinear image processing to two inverse imaging problems. The first problem involves image restoration, where the aim is to reconstruct an unknown high-quality image from a low-quality observed image. Sparse representations of images have drawn a considerable amount of interest in recent years. The assumption that natural signals, such as images, admit a sparse decomposition over a redundant dictionary leads to efficient algorithms for handling such sources of data. The standard sparse representation, however, does not consider the intrinsic geometric structure present in the data, thereby leading to sub-optimal results. Using the concept that a signal is block sparse in a given basis —i.e., the non-zero elements occur in clusters of varying sizes — we present a novel and efficient algorithm for learning a sparse representation of natural images, called graph regularized block sparse dictionary (GRBSD) learning. We apply the proposed method towards two image restoration applications: 1) single-Image super-resolution, where we propose a local regression model that uses learned dictionaries from the GRBSD algorithm for super-resolving a low-resolution image without any external training images, and 2) image inpainting, where we use GRBSD algorithm to learn a multiscale dictionary to generate visually plausible pixels to fill missing regions in an image. Experimental results validate the performance of the GRBSD learning algorithm for single-image super-resolution and image inpainting applications. The second problem addressed in this work involves image enhancement for detection and segmentation of objects in images. We exploit the concept that even though data from various imaging modalities have high dimensionality, the data is sufficiently well described using low-dimensional geometrical structures. To facilitate the extraction of objects having such structure, we have developed general structure enhancement methods that can be used to detect and segment various curvilinear structures in images across different applications. We use the proposed method to detect and segment objects of different size and shape in three applications: 1) segmentation of lamina cribrosa microstructure in the eye from second-harmonic generation microscopy images, 2) detection and segmentation of primary cilia in confocal microscopy images, and 3) detection and segmentation of vehicles in wide-area aerial imagery. Quantitative and qualitative results show that the proposed methods provide improved detection and segmentation accuracy and computational efficiency compared to other recent algorithms.