Orientadores: Rafael Santos Mendes, Jonatas Manzolli / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-10T02:37:04Z (GMT). No. of bitstreams: 1 Luvizotto_AndreLuiz_M.pdf: 5323411 bytes, checksum: 87b1961bd111c6cc93b0d048c33bd164 (MD5) Previous issue date: 2007 / Resumo: Neste trabalho, as transformadas wavelet são utilizadas para recodificar e explorar sinais musicais. A idéia central do trabalho é mostrar que a reordenação adequada dos coeficientes provenientes das análises possibilita não somente recodificar de modo econômico o sinal como também transitar por sonoridades distintas. Ou seja, relaciona-se o posicionamento dos coeficientes wavelets, dentro de cada nível de multiresolução, com a geração de diferentes timbres. Uma vez definido o ordenamento dos coeficientes, duas formas de determinação de seus valores são exploradas. Na primeira forma, essas curvas ascendentes de coeficientes são aproximadas por uma função polinomial de grau p, levando a representações econômicas do sinal musical. Já no segundo modelo, os coeficientes de um sinal distinto (chamado sinal de base) são ordenados de forma ascendente. São então reordenados através das posições equivalentes do sinal original, também chamado de alvo, permitindo a sua reobtenção e também o trânsito entre as duas sonoridades envolvidas. Os resultados deste trabalho destacam a grande importância que o posicionamento dos coeficientes exercem na sonoridade com relação aos seus valores. Com os experimentos realizados, foi possível constatar que valores aproximados de coeficientes, corretamente dispostos no tempo, geram timbres alvos diversos de forma satisfatória. As análises dos resultados são feitas por metodologias diversas e pela audição dos arquivos de áudio gerados, que acompanham o texto em cd. Por fim, uma discussão sobre os resultados obtidos é realizada e uma proposta de continuação da pesquisa é sugerida, baseada em grupos de permutações como forma de síntese / Abstract: Musical signals are encoded and represented using wavelet transforms. Starting upon wavelet analysis, the main research idea is to show how an adequate re-ordination of wavelet coefticients makes possible to decode the signal economically and also to obtain different sonorities. Moreover, we relate a specific set of wavelet coeflicients from °each multi-resolution leveI to generate new timbres. Given re-ordination procedures, two forms to determine their values are explored. First one, coeflicient curves in ascendent order are approximated by a polynomial function of degree p that leads to an economic representation of musical signals. In the second approach, coeflicients from another signal (named "base"signal) are ordinated in ascendent order also. After they are re-ordinate across the positions ofthe original signal, called "target"signal. The results ofthis research highlight the great importance of the wavelet coeflicient's order to manipulate timbre. Experiments presented . here showed that coeflicients with approximated values and correct1y disposed in time, can be used to generate target timbres in a satisfactory way. The analysis of the results were done by different methodologies and by listening to the sound examples from the CD attached to this dissertation. Finally, a discussion of research aims is presented and a proposal for further work based on Mathematical Group of Permutation is projected as a way to develop a new synthesis method / Mestrado / Engenharia de Computação / Mestre em Engenharia Elétrica

The last few decades have witnessed what is commonly known as an "information explosion". The advent of the Internet has enabled the sharing of large amounts of information between users in almost every part of the world. This had led to increased bandwidth and storage requirements. These requirements necessitate the development of new techniques to optimize the utilization of existing bandwidth and to minimize storage requirements. In emerging areas of real-time multimedia applications, researchers have successfully developed new approaches, employing techniques such as transform coding, prediction, vector quantization, etc, to efficiently represent and consequently compress data, especially image data. Recently, multiple transform domain representation techniques have been reported which successfully compress one and multidimensional signals. In this thesis, a novel and efficient codec using multiple transform domain representation in conjunction with split vector quantization is presented, which provides superior coding performance for images. An adaptive scheme that further enhances the representation accuracy of the above codec is also described. Results of simulations of the proposed technique are presented that confirm the improved performance of the codec.

Thesis (Ph. D.)--University of California, Riverside, 2009. / Includes abstract. Title from first page of PDF file (viewed January 27, 2010). Includes bibliographical references (p. 99-105). Issued in print and online. Available via ProQuest Digital Dissertations.

Thesis (Ph. D.)--Mississippi State University. Department of Electrical and Computer Engineering. / Title from title screen. Includes bibliographical references.

Thesis (Ph.D.)--University of Wollongong, 2006. / Typescript. Includes bibliographical references: leaf 236-258.

Thesis (Ph. D.)--State University of New York at Binghamton, Thomas J. Watson School of Engineering and Applied Science, Department of Electrical and Computer Engineering, 2008. / Includes bibliographical references.

Thesis (Ph. D.)--Duke University, 2007. / Includes bibliographical references.

Thesis (Ph.D.) -- McMaster University, 2001. / Includes bibliographical references (leaves 120-128). Also available via World Wide Web.

Thesis (Ph. D.)--Ohio State University, 2005. / Title from first page of PDF file. Document formatted into pages; contains i-xxii, xx-xxi, 183 p.; also includes graphics. Includes bibliographical references (p. 171-183). Available online via OhioLINK's ETD Center

The Blind Image Deconvolution/Deblurring (BID) problem was realised in the early 1960s but it still remains a challenging task for the image processing research community to find an efficient, reliable and most importantly a diversely applicable deblurring scheme. The main challenge arises from little or no prior information about the image or the blurring process as well as the lack of optimal restoration filters to reduce or completely eliminate the blurring effect. Moreover, restoration can be marred by the two common side effects of deblurring; namely the noise amplification and ringing artefacts that arise in the deblurred image due to an unrealizable or imperfect restoration filter. Also, developing a scheme that can process different types of blur, especially for real images, is yet to be realized to a satisfactory level. This research is focused on the development of blind restoration schemes for real life blurred images. The primary objective is to design a BID scheme that is robust in term of Point Spread Function (PSF) estimation, efficient in terms of restoration speed, and effective in terms of restoration quality. A desired scheme will require a deblurring measure to act as a feedback of quality regarding the deblurred image and lead the estimation of the blurring PSF. The blurred image and the estimated PSF can then be passed on to any classical restoration filter for deblurring. The deblurring measures presented in this research include blind non-Gaussianity measures as well as blind Image Quality Measures (IQMs). These measures are blind in the sense that they are able to gauge the quality of an image directly from it without the need to reference a high quality image. The non-Gaussianity measures include spatial and spectral kurtosis measures; while the image quality analysers include the Blind/Reference-less Image Spatial QUality Evaluator (BRISQUE), Natural Image Quality Evaluator (NIQE) index and Reblurring based Peak Signal to Noise Ratio (RPSNR) measure. BRISQUE, NIQE and spectral kurtosis, are introduced for the first time as deblurring measures for BID. RPSNR is a novel full reference yet blind IQM designed and used in this research work. Experiments were conducted on different image datasets and real life blurred images. Optimization of the BID schemes has been achieved using a gradient descent based scheme and a Genetic Algorithm (GA). Quantitative results based on full-reference and non-reference IQMs, present BRISQUE as a robust and computationally efficient blind feedback quality measure. Also, parametric and arbitrarily shaped (non-parametric or generic) PSFs were treated for the blind deconvolution of images. The parametric forms of PSF include uniform Gaussian, motion and out-of-focus blur. The arbitrarily shaped PSFs comprise blurs that have a much more complex blur shape which cannot be easily modelled in the parametric form. A novel scheme for arbitrarily shaped PSF estimation and blind deblurring has been designed, implemented and tested on artificial and real life blurred images. The scheme provides a unified base for the estimation of both parametric and arbitrarily shaped PSFs with the BRISQUE quality measure in conjunction with a GA. Full-reference and non-reference IQMs have been utilised to gauge the quality of deblurred images for the BID schemes. In the real BID case, only non-reference IQMs can be employed due to the unavailability of the reference high quality image. Quantitative results of these images depict the restoration ability of the BID scheme. The significance of the research work lies in the BID scheme‘s ability to handle parametric and arbitrarily shaped PSFs using a single algorithm, for single-shot blurred images, with enhanced optimization through the gradient descent scheme and GA in conjunction with multiple feedback IQMs.