In this thesis, we proposed two frequency-based watermarking algorithms. One is DCT-based method. Embedding watermark in the multi-areas and multi- frequency bands to ensure we can get a less distorted watermark sequence under unintentional circumstance. The other is DWT-based method. The parent-children relationship and the feature of bit-plane coding in the EZW algorithm are exploited to embed watermark. It makes that we can know the watermark exist or not in the progressive transmission system. The experimental results show that the proposed methods both can resist unintentional attacks. The DWT-based method also has a better progressive detection capability.

none

The presence of subsurface anomalies, such as cavities, faults, unknown tunnels, etc., either natural or man-made, can cause public safety hazards. The detection of these features requires the development of new methods. Seismic Rayleigh surface wave imaging is a relatively new non-destructive testing technique (NDT) which generates subsurface images without drilling boreholes into the ground, and in recent years has been widely used for soil characterization in geotechnical investigations. In the last decade, some researchers have applied the technique to near-surface imaging and showed the possibility and potential for engineering applications. This research presents the development of a technique to process seismic Rayleigh waves to detect and image subsurface anomalies. This study conducted investigations of Rayleigh wave behaviors and developed a new strategy for Rayleigh wave isolation from raw field data. The strategy applies wavelet transforms, instead of the conventional spectral analysis of surface waves (SASW) method, or popular multichannel analysis of surface waves (MASW) techniques, to pair-channel analysis of the isolated Rayleigh wave data for dispersion calculation. Finally, a simple steady inversion technique was applied to yield shear velocity as a function of both depth and distance, and shear velocity field images (SVF), for near surface section display. This research consists of development, computer programming, field tests, data processing and interpretation. Three sites in different scenarios were used for seismic investigations: old mining tunnels in medium dipping coal seams in Stellarton coalfield, mining cavities in steeply dipping gold-bearing veins in West Waverley Gold District and an anomaly in nearly horizontal strata in Liverpool. All these sites are in the province of Nova Scotia, Canada. The results from seismic surface wave technique introduced in this research can be evaluated by field observations, documents and borehole logs. The satisfactory interpretations and success of this investigation shows that this technique is suitable for engineering application for subsurface investigations.

A 3D wavelet based encoding spectroscopic method (WE-SI) is investigated and implemented on a 3 Tesla Siemens Scanner. Compared to CSI, the proposed method is able to reduce acquisition time, and preserves the spatial metabolite distribution. As expected, a decrease in Signal to Noise Ratio (SNR) is noticed in WE-SI data compared to CSI. The dissertation explores important physical principles in MRI and spectroscopic imaging as a background, following by introduction of the wavelet encoding theory and comparison to Fourier encoding.

A 3D wavelet based encoding spectroscopic method (WE-SI) is investigated and implemented on a 3 Tesla Siemens Scanner. Compared to CSI, the proposed method is able to reduce acquisition time, and preserves the spatial metabolite distribution. As expected, a decrease in Signal to Noise Ratio (SNR) is noticed in WE-SI data compared to CSI. The dissertation explores important physical principles in MRI and spectroscopic imaging as a background, following by introduction of the wavelet encoding theory and comparison to Fourier encoding.

This thesis explains and tests a wavelet based implicit numerical method for the solving of partial differential equations. Intended for problems with localized small-scale interactions, the method exploits the form of the wavelet decomposition to divide the implicit system created by the time discretization into multiple, smaller, systems that can be solved sequentially. Included are tests of this method on linear and non-linear problems, with both its results and the time required to calculate them compared to basic models. It was found that the method requires less computational effort than the high resolution control results. Furthermore, the method showed convergence towards high resolution control results.

Chemnitz, Techn. Univ., Diplomarb., 2007.

Zugl.: Aachen, Techn. Hochsch., Diss., 2009

Zugl.: Chemnitz, Techn. Univ., Diss., 2007 / Hergestellt on demand

As bolsas para coleta de sangue necessitam cuidados técnicos que assegurem sua esterilidade biológica durante o processo de fabricação até a entrega do produto final. As bolsas para coleta de sangue não devem conter qualquer tipo de resíduo, para que o seu uso ou conteúdo não seja comprometido causando diversos transtornos. É preciso eliminar o risco de contaminação; no entanto muitas empresas fazem a inspeção de bolsas para coleta de sangue visualmente, a olho nu, por funcionários treinados no uso de equipamento simples, o qual não pode garantir que todas as bolsas estejam livres de resíduos, por depender do fator humano. Para assegurar a esterilização das bolsas comercializadas, este trabalho propõe uma metodologia para automação de um sistema de visão computacional baseado em conjunto de técnica de extração de características de imagens e reconhecimento de padrões para detectar resíduos em imagens de bolsas para coleta de sangue durante o processo de fabricação, e assim, melhorar a qualidade das bolsas comercializadas. Os resíduos encontrados nas imagens capturadas se apresentam como componentes de alta freqüência. A metodologia proposta utiliza a Transformada Wavelet (Wavelet de Haar) em dois níveis de decomposição, como filtro passa - alta para destacar as altas freqüências, e assim detectar a existência de resíduos nas imagens capturadas. Os resultados obtidos com a inspeção visual manual realizada em bolsas para coleta de sangue, mostram que existem resultados divergentes na análise de uma mesma bolsa, o que reforça a necessidade de automatização dessa tarefa para que sejam obtidos resultados mais uniformes com padrões definidos, e dessa maneira melhorar a qualidade das bolsas comercializadas para coleta de sangue. / The blood bags require technician attention to assure its biological sterility in the process of manufacture and in the final product. The blood bags must not contain any type of waste, so that its use or content is not compromised causing various disorders; it is necessary to eliminate any contamination risk; however many companies perform a visually inspection of the blood bags, by naked eye, by trained employees in the use of a simple equipment, which cannot guarantee that all the bags are residue-free because depending on the human factor. The residues showed on the pictures were characterized as high frequency components. So the proposed methodology applies two levels decomposition of Haar´s Wavelet as high pass filter to evidence high frequencies, and to evidence the residues at the pictures and to evidence the existence of residues on the captured pictures. To assure the sterilization of the commercialized blood bags, this work propose a methodology of an automated computational system based in a group of techniques of image characteristics extraction and recognition patterns to detect residues in blood bag images in the process of manufacture, in order to improve the quality of the commercialized blood bags.This technique is used as high-pass filter, to detach the high frequencies, using a Haar-wavelet in two decomposition levels. The results obtained with the manual visual inspection indicate that were found divergent results in the analysis of the same blood bag, which reinforces the necessity of the automation of this task in order to obtain more uniform results with defined standards.