A codificação distribuída de vídeo constitui um novo paradigma em compressão de vídeo frente aos codificadores híbridos da família MPEG-x e H.26x. Nesses codificadores, a estimação de movimento é a principal etapa do processo de compressão do sinal de vídeo. Desta forma, a codificação demanda um alto custo computacional exigindo desempenho do codificador. Neste trabalho é apresentado um codec de vídeo baseado na teoria da codificação distribuída com perdas. O compressor apresentado codifica os quadros ímpares e os quadros pares separadamente utilizando códigos turbo e a transformada wavelet. O processo de decodificação é feito de forma iterativa e explora a dependência estatística entre os quadros da seqüência de vídeo original. Esta abordagem permite uma redução bastante significativa no tempo de processamento envolvido na compressão do sinal de vídeo, tornando viável a implementação deste tipo de codificador em dispositivos com recursos escassos de processamento e memória. Os resultados obtidos em simulações comprovam o bom desempenho do codec proposto em relação ao padrão estado da arte em compressão de vídeo, o H.264/AVC; Distributed video coding is a new paradigm for video compression in opposition over the existing video coding standards like MPEG-x and H.26x families. These codecs make use of motion estimation algorithms...

Wavelet transforms provide a new technique for time-scale analysis of non-stationary signals. Wavelet analysis uses orthonormal bases in which computations can be done efficiently with multirate systems known as filter banks. This thesis develops a comprehensive set of tools for (multidimensional) multirate signal analysis and uses them to investigate two multirate systems: filter banks and transmultiplexers. Several results in filter bank theory are obtained: a new parameterization of unitary filter banks, a theory of modulated filter banks, a theory of filter banks with symmetry restrictions, reduction of the multidimensional rational sampling rate filter bank problem to the uniform sampling rate filter bank problem, solution to the completion problem for filter banks (by reducing it to the (YJBK) parameterization problem in control theory) etc. Perfect reconstruction filter banks are shown to give structured decompostions of separable Hilbert spaces. Filter banks are used to construct several classes of wavelet bases: multiplicity M wavelet tight frames and frames, regular multiplicity M orthonormal bases, modulated wavelet tight frames etc. The thesis describes the design of optimal wavelets for signal representation and the wavelet sampling theorem. Application of wavelets in signal interpolation and in the approximation of linear-translation invariant operators is investigated.

O eletrocardiograma (ECG) é uma ferramenta utilizada para o diagnóstico de cardiopatias e outras doenças. Este trabalho tem como objetivo a detecção do complexo QRS, com foco na onda R, que representa a contração dos ventrículos. Para isso, são apresentadas duas técnicas de processamento do sinal de ECG. A primeira utiliza o algoritmo proposto por Pan & Tompkins que consiste em um banco de filtros digitais. A segunda faz uso da transformada wavelet discreta, que permite a localização de características de sinais tanto no tempo quanto na frequência. É apresentado um comparativo da eficácia dos dois algoritmos com base na sua implementação através de FPGA, utilizando dois métodos, o processamento serial em microcontrolador programado em C e o paralelo inteiramente em VHDL, com o intuito de comparar os tempos de processamento. Os resultados sugerem que trabalhos futuros poderão ser baseados na investigação de outras famílias wavelets para a detecção do complexo QRS em sinais de ECG, bem como explorar outros métodos de implementação de filtros em FPGA; The electrocardiogram (ECG) is a tool used for diagnosis of diseases related to the heart. This work has the purpose of detecting QRS complex, focusing on the R wave...

IEEE-standard, iterative and direct linear system solution methods, eigendecomposition and model-order reduction, fast Fourier transforms, multigrid, wavelets and other multiresolution methods, matrix sparsification. Nonlinear root finding (Newton's method). Numerical interpolation and extrapolation. Quadrature.

We construct two new families of wavelets: One family of frames which is well suited for frequency localized signals and interpolates between the standard wavelet frames and a version of a Gabor type frame. The second family is well suited for time localized signals and interpolated between a version of a wavelet frame and a standard Gabor frame. In particular we approximate Gabor analysis by wavelets. Our construction is based on certain realizations of the unitary representations of the Heisenberg group and of the affine group on L^2(R). The main technical tool that we use for the interpolation procedures is contraction of Lie groups representations.

We introduce a new family of multivariate wavelets which are obtained by "polyharmonic subdivision". They generalize directly the original compactly supported Daubechies wavelets.; Comment: 6 pages, prepared for the ACM proceedings of CompSysTech 2010

In this paper we present a general approach to multivariate periodic wavelets generated by scaling functions of de la Vall\'ee Poussin type. These scaling functions and their corresponding wavelets are determined by their Fourier coefficients, which are sample values of a function, that can be chosen arbitrarily smooth, even with different smoothness in each direction. This construction generalizes the one-dimensional de la Vall\'ee Poussin means to the multivariate case and enables the construction of wavelet systems, where the set of dilation matrices for the two-scale relation of two spaces of the multiresolution analysis may contain shear and rotation matrices. It further enables the functions contained in each of the function spaces from the corresponding series of scaling spaces to have a certain direction or set of directions as their focus, which is illustrated by detecting jumps of certain directional derivatives of higher order.; Comment: 29 pages, 18 figures

New elliptic cylindrical wavelets are introduced, which exploit the relationship between analysing filters and Floquet's solution of Mathieu differential equations. It is shown that the transfer function of both multiresolution filters is related to the solution of a Mathieu equation of odd characteristic exponent. The number of notches of these analysing filters can be easily designed. Wavelets derived by this method have potential application in the fields of optics, microwaves and electromagnetism.; Comment: 5 pages, 4 figures. in: 2002 WSEAS International Conference on Wavelet Analysis and Multirate Systems, Vouliagmeni, Greece. arXiv admin note: substantial text overlap with arXiv:1501.07255

We develop a general notion of orthogonal wavelets `centered' on an irregular knot sequence. We present two families of orthogonal wavelets that are continuous and piecewise polynomial. We develop efficient algorithms to implement these schemes and apply them to a data set extracted from an ocelot image. As another application, we construct continuous, piecewise quadratic, orthogonal wavelet bases on the quasi-crystal lattice consisting of the $\tau$-integers where $\tau$ is the golden ratio. The resulting spaces then generate a multiresolution analysis of $L^2(\mathbf{R})$ with scaling factor $\tau$.; Comment: 37 pages, 9 figures

In recent years, a rapidly growing literature has focussed on the construction of wavelet systems to analyze functions defined on the sphere. Our purpose in this paper is to generalize these constructions to situations where sections of line bundles, rather than ordinary scalar-valued functions, are considered. In particular, we propose {\em needlet-type spin wavelets} as an extension of the needlet approach recently introduced by Narcowich, Petrushev and Ward, and then considered for more general manifolds by Geller and Mayeli. We discuss localization properties in the real and harmonic domains, and investigate stochastic properties for the analysis of spin random fields. Our results are strongly motivated by cosmological applications, in particular in connection to the analysis of Cosmic Microwave Background polarization data.; Comment: 37 pages

We consider the application of the polyharmonic subdivision wavelets (of Daubechies type) to Image Processing, in particular to Astronomical Images. The results show an essential advantage over some standard multivariate wavelets and a potential for better compression.; Comment: 9 pages

Numerical optimization is used to construct new orthonormal compactly supported wavelets with Sobolev regularity exponent as high as possible among those mother wavelets with a fixed support length and a fixed number of vanishing moments. The increased regularity is obtained by optimizing the locations of the roots the scaling filter has on the interval (pi/2,\pi). The results improve those obtained by I. Daubechies [Comm. Pure Appl. Math. 41 (1988), 909-996], H. Volkmer [SIAM J. Math. Anal. 26 (1995), 1075-1087], and P. G. Lemarie-Rieusset and E. Zahrouni [Appl. Comput. Harmon. Anal. 5 (1998), 92-105].; Comment: 18 pages, 8 figures

We define an affine structure on $\ltwor\oplus...\oplus\ltwor$ and, following some ideas developed in \cite{Dut1}, we construct a local trace function for this situation. This trace function is a complete invariant for a shift invariant subspace and it has a variety of properties which make it easily computable. The local trace is then used to give a characterization of super-wavelets and to analyze their multiplicity function, dimension function and spectral function. The "$n\times$" oversampling result of Chui and Shi \cite{CS} is refined to produce super-wavelets.

In any quasi-metric space of homogeneous type, Auscher and Hyt\"onen recently gave a construction of orthonormal wavelets with H\"older-continuity exponent $\eta>0$. However, even in a metric space, their exponent is in general quite small. In this paper, we show that the H\"older-exponent can be taken arbitrarily close to 1 in a metric space. We do so by revisiting and improving the underlying construction of random dyadic cubes, which also has other applications.

Adapting the recently developed randomized dyadic structures, we introduce the notion of spline function in geometrically doubling quasi-metric spaces. Such functions have interpolation and reproducing properties as the linear splines in Euclidean spaces. They also have H\"older regularity. This is used to build an orthonormal basis of H\"older-continuous wavelets with exponential decay in any space of homogeneous type. As in the classical theory, wavelet bases provide a universal Calder\'on reproducing formula to study and develop function space theory and singular integrals. We discuss the examples of $L^p$ spaces, BMO and apply this to a proof of the T(1) theorem. As no extra condition {(like 'reverse doubling', 'small boundary' of balls, etc.)} on the space of homogeneous type is required, our results extend a long line of works on the subject.; Comment: We have made improvements to section 2 following the referees suggestions. In particular, it now contains full proof of formerly Theorem 2.7 instead of sending back to earlier works, which makes the construction of splines self-contained. One reference added

A classical theorem attributed to Naimark states that, given a Parseval frame $\mathcal{B}$ in a Hilbert space $\mathcal{H}$, one can embed $\mathcal{H}$ in a larger Hilbert space $\mathcal{K}$ so that the image of $\mathcal{B}$ is the projection of an orthonormal basis for $\mathcal{K}$. In the present work, we revisit the notion of Parseval frame MRA wavelets from two papers of Paluszy\'nski, \v{S}iki\'c, Weiss, and Xiao (PSWX) and produce an analog of Naimark's theorem for these wavelets at the level of their scaling functions. We aim to make this discussion as self-contained as possible and provide a different point of view on Parseval frame MRA wavelets than that of PSWX.; Comment: 19 pages

We apply successfully the Compressive Sensing approach for Image Analysis using the new family of Polyharmonic Subdivision wavelets. We show that this approach provides a very efficient recovery of the images based on fewer samples than the traditional Shannon-Nyquist paradigm. We provide the results of experiments with PHSD wavelets and Daubechies wavelets, for the Lena image and astronomical images.; Comment: 11 pages, 10 figures

We propose a novel method for constructing wavelet transforms of functions defined on the vertices of an arbitrary finite weighted graph. Our approach is based on defining scaling using the the graph analogue of the Fourier domain, namely the spectral decomposition of the discrete graph Laplacian $\L$. Given a wavelet generating kernel $g$ and a scale parameter $t$, we define the scaled wavelet operator $T_g^t = g(t\L)$. The spectral graph wavelets are then formed by localizing this operator by applying it to an indicator function. Subject to an admissibility condition on $g$, this procedure defines an invertible transform. We explore the localization properties of the wavelets in the limit of fine scales. Additionally, we present a fast Chebyshev polynomial approximation algorithm for computing the transform that avoids the need for diagonalizing $\L$. We highlight potential applications of the transform through examples of wavelets on graphs corresponding to a variety of different problem domains.

Wavelets have been shown to be effective bases for many classes of natural signals and images. Standard wavelet bases have the entire vector space $\mathbb R^n$ as their natural domain. It is fairly straightforward to adapt these to rectangular subdomains, and there also exist constructions for domains with more complex boundaries. However those methods are ineffective when we deal with domains that are very arbitrary and convoluted. A particular example of interest is the human cortex, which is the part of the human brain where all the cognitive activity takes place. In this thesis, we use the lifting scheme to design wavelets on arbitrary volumes, and in particular on volumes having the structure of the human cortex. These wavelets have an element of randomness in their construction, which allows us to repeat the analysis with many different realizations of the wavelet bases and averaging the results, a method that improves the power of the analysis. Next, we apply this type of wavelet transforms to the statistical analysis to fMRI data, and we show that it enables us to achieve greater spatial localization than other, more standard techniques.; Comment: PhD Dissertation, Princeton University, PACM, Adviser: Ingrid Daubechies, September 2012