A presente tese descreve a construção de uma nova transformada wavelet, aqui chamada de SPIKELET, que, combinada com um algoritmo proposto, é aplicada no reconhecimento computacional de padrões em spikes (picos) e spikes sobrepostos (overlaps) encontrados no sinal digitalizado correspondente às reações do neurônio H1 do cérebro de uma mosca de ordem Diptera, que é sensível aos estímulos visuais do meio externo. O algoritmo fornece, além do formato do sinal encontrado, o \'\'instante\'\' em que ele ocorreu, sendo que a implementação é feita, inclusive, em tempo-real, com o uso de um DSP. / This thesis describes the construction of a new wavelet transform, that is called SPIKELET, which is used together with a proposed algorithm, for spikes and overlaps pattern recognition, in a digitalized signal corresponding to the H1 visual neuron action potential from a Diptera\'s fly brain. The algorithm provides both the shape of the identified signal and the \'\'instant\'\' of time it happened. The implementation is also done in real time, using a DSP.

A presente tese descreve a construção de uma nova transformada wavelet, aqui chamada de SPIKELET, que, combinada com um algoritmo proposto, é aplicada no reconhecimento computacional de padrões em spikes (picos) e spikes sobrepostos (overlaps) encontrados no sinal digitalizado correspondente às reações do neurônio H1 do cérebro de uma mosca de ordem Diptera, que é sensível aos estímulos visuais do meio externo. O algoritmo fornece, além do formato do sinal encontrado, o \'\'instante\'\' em que ele ocorreu, sendo que a implementação é feita, inclusive, em tempo-real, com o uso de um DSP. / This thesis describes the construction of a new wavelet transform, that is called SPIKELET, which is used together with a proposed algorithm, for spikes and overlaps pattern recognition, in a digitalized signal corresponding to the H1 visual neuron action potential from a Diptera\'s fly brain. The algorithm provides both the shape of the identified signal and the \'\'instant\'\' of time it happened. The implementation is also done in real time, using a DSP.

No description available.

Yes / A new wavelet transform (WT) is introduced based on the fractional properties of the traditional Fourier transform. The new wavelet follows from the fractional Fourier order which uniquely identifies the representation of an input function in a fractional domain. It exploits the combined advantages of WT and fractional Fourier transform (FrFT). The transform permits the identification of a transformed function based on the fractional rotation in time-frequency plane. The fractional rotation is then used to identify individual fractional daughter wavelets. This study is, for convenience, limited to one-dimension. Approach for discussing two or more dimensions is shown.

The High Resolution Electrocardiogram (HRECG) is used to detect Ventricular Late Potentials (VLPs) in post-myocardial infarction patients. VLPs are low-amplitude, high-frequency signals that are usually found within the terminal part of the QRS complex. The aim of this research was to develop possible alternative methods and improve existing methods of detecting VLP activity. There are two main topics in this work: applications of the Continuous Wavelet Transform (CWT) and the Discrete Wavelet Transform (DWT) to the HRECG. For the CWT application, a Fractionation Factor (FF) method proposed by previous work was further investigated and improved by combining the CWT and DWT for distinction between patients with and those without VLPs. A Differential Fractionation Factor was proposed as an alternative approach to the FF with better results. Observation in the time-scale plot showed a difference in the energy distribution. A 2-dimensional Fractionation Factor was proposed to quantify this difference. A new concept of local intermittency was investigated to exhibit energy nonuniformity and then a Local Intermittency Factor was developed to quantify the degree of nonuniformity. The energy computed with the CWT was also used for patient distinction. Patients with VLPs may be also characterised by a slow rate of energy decay. The CWT can reveal a difference in ECG irregularity between the patients. A new approach of approximate entropy was implemented to quantify this irregularity. For the DWT application, the DWT can reveal irregularity of VLP activity and it was quantified by the approximate entropy to identify patients with VLPs. The wavelet entropy was utilised as an alternative method to the FF. The energy computed with the DWT was used for patient classification. The potentially promising results of both the CWT and DWT applications were obtained from the methods of computing the energy and approximate entropy

No description available.

No description available.

No description available.

No description available.

The digital watermarking technique embeds meaningful information into one or more watermark images hidden in one image, in which it is known as a secret carrier. It is difficult for a hacker to extract or remove any hidden watermark from an image, and especially to crack so called digital watermark. The combination of digital watermarking technique and traditional image encryption technique is able to greatly improve anti-hacking capability, which suggests it is a good method for keeping the integrity of the original image. The research works contained in this thesis include: (1)A literature review the hyperchaotic watermarking technique is relatively more advantageous, and becomes the main subject in this programme. (2)The theoretical foundation of watermarking technologies, including the human visual system (HVS), the colour space transform, discrete wavelet transform (DWT), the main watermark embedding algorithms, and the mainstream methods for improving watermark robustness and for evaluating watermark embedding performance. (3) The devised hyperchaotic scrambling technique it has been applied to colour image watermark that helps to improve the image encryption and anti-cracking capabilities. The experiments in this research prove the robustness and some other advantages of the invented technique. This thesis focuses on combining the chaotic scrambling and wavelet watermark embedding to achieve a hyperchaotic digital watermark to encrypt digital products, with the human visual system (HVS) and other factors taken into account. This research is of significant importance and has industrial application value.