[Home] [Back Issues] [Search] [Vol 1] [Vol 2] [Vol 3] [Vol 4] [Vol 5]

The Wavelet Digest - Volume 3, Issue 12 (August 5, 1994)


Today's Editor: Wim Sweldens
                wim.sweldens@cs.kuleuven.ac.be


Today's Topics:
     0. Announce: Wavelet Digest joins the World Wide Web
     1. Question: Using the wavelet transform to analyze complex structure
     2. Question: Matrix Compression
     3. Preprint: Some Remarks About Compactly Supported Spline Wavelets
     4. Preprint: Two preprints from B. Rubin available
     5. Question: Wavelets and time series
     6. Preprint: (p)reprints from X. Xia and co-workers available
     7. Question: Applications of wavelets to spectrum analysis
     8. General:  JPEG type quantization matrices for Wavelet Transforms
     9. Question: Contributions to library of algorithms
    10. Book:     Wavelets in geophysics
    11. Contents: Constructive Approximation vol 10 nr 3
    12. Book:     Geometry-Driven Diffusion in Computer Vision

Submissions:
  E-mail to wavelet@math.sc.edu with "submit" as subject.

Subscriptions:
  E-mail to wavelet@math.sc.edu with "subscribe" as subject.
  To unsubscribe, e-mail with "unsubscribe" followed by your e-mail
  address as subject. To change address, unsubscribe and resubscribe.

Preprints, references, and back issues can be obtained from our
information servers:
  Ftp:    ftp.math.sc.edu (/pub/wavelet)
  Gopher: gopher.math.sc.edu
  WWW:    http://www.math.sc.edu/~wavelet/


Current number of subscribers: 3575

Calender of events:

 Aug  17-20: Wavelets for Engineering Applications, College Station WD 3.7  #12
 Sep  12-16: UCLA Short Course on Wavelets                          WD 3.11 #10
 Sep  25-28: Visual Communications and Image Processing, Chicago    WD 2.16 #3
 Sep  26-30: Total Positivity and its Applications, Jaca, Spain     WD 3.2  #8
 Oct  25-28: IEEE Symposium on Time-Frequency Analysis, Philadelph. WD 3.6  #11
 Nov  1-2:   Wavelets in Medicine and Biology, Baltimore, Maryland  WD 2.17 #11
 Nov  13-16: Artificial Neural Networks in Engineering, St. Louis   WD 3.4  #8
 Nov  13-16: IEEE international conference on image processing      WD 3.9  #11
 Nov  16-18: Meeting on wavelets and statitics, Grenoble, France    WD 3.5  #3
 Dec  16-19: Methods and Applications of Analysis, Hong Kong        WD 3.10 #7
 --- 1995 ---
 Jan  8-12:  Approximation Theory, College Station                  WD 3.7  #4
 

From: Editor Subject: Announce: Wavelet Digest joins the World Wide Web Dear Wavelet-Digest-reader, Since recently, the Wavelet Digest has joined the rapidly growing World Wide Web (WWW). The URL of the home page is http://www.math.sc.edu/~wavelet/. The information present is essentially the same as available through our ftp site. Special thanks to Geert Uytterhoeven (uytterho@math.sc.edu) for putting the pages together. Please let us know of any questions, comments, or problems. Also, we welcome other links to wavelet related information on the WWW.
From: Jean Bragard <jbragard@pluri.ucm.es> Subject: Question: Using the wavelet transform to analyze complex structure Dear wavelet-friends, I'm interested in using wavelet transform to analyze complex structure in order to measure the level of defects. To be specific, a perfect hexagonal pattern is perturbed by an external constraint. I would like to relate the number of defects to the constraint using the wavelet transform. Could you give me some hints in order to tackle the problem efficiently and is there any program directly available that will do a WT of a 2D pattern ? Thanks in advance. Jean BRAGARD Instituto Pluridisciplinar Complutense Unidad de Fluidos Madrid
From: wvl@ipp-garching.mpg.de (Wolfgang von der Linden) Subject: Question: Matrix Compression Dear Colleagues, is there anybody around, who knows how to compress a matrix of the form M_{ij} = e^{a * ((x_i-x_j)^2 + (y_i-y_j)^2)} using discrete wavelet transform. Where (x_i,y_i) are point on a 2d square lattice, e.g. x_i = mod(i,256) y_i = integer(i/256) The approach, given in Numerical Recipes, gives merely a compression by 50% if one discards matrix elements of size 10^{-6}, which is worse than the original matrix and in disagreement to W.Press's statement that the number of matrix elements, larger than eps, is 10*N*Log_{10}(eps) Thanks Wolfgang von der Linden My e-mail address reads (wvl@ibmop5.ipp-garching.mpg.de)
From: eduser@mate.dm.uba.ar (Eduardo Pedro Serrano) Subject: Preprint: Some Remarks About Compactly Supported Spline Wavelets Preprint : 'Some Remarks About Compactly Supported Spline Wavelets' (Almost orthogonal wavelets with compact support) Abstract: In this paper we propose an extended family of almost orthogonal spline wavelets with compact support. These functions provide snug bases for L2(R), preserving semiorthonormal properties. As it is well known, while orthogonality gives us non redundant representations, finite support has attractive features for numerical implementations. This work represents an effort to combine both desirable conditions in the spline case and to enhance previous results of Ch. Chui and Unser et al. We star by reviewing the concept of semiorthogonal wavelets and we discuss its performance. Next, we give a brief description of the general technique for computing compactly supported spline wavelets, and at last we expose the above mentioned functions. Also we develop several formulas in accord with our purposes. Author: Eduardo Serrano Ad: Dep Mat. FCEyN - Universidad de Buenos Aires - Argentina E-mail:eduser@mate.dm.uba.ar For more information please contact us. Thanks in advance for any comments.
From: boris@humus.huji.ac.il Subject: Preprint: Two preprints from B. Rubin available The following preprints of B.Rubin are available by anonimous ftp from sunset.ma.huji.ac.il. name: ftp Password: boris@sunset.huji.ac.il Directory: rubin filename: wtr.tex CONTINUOUS WAVELET TRANSFORMS ON A SPHERE A general method for constructing continuous wavelet transforms is suggested. Such transforms arise naturally as a result of special procedure of analitycal continuation for analytic families of operators. As an example the case of n-dimensional sphere in $R^{n+1}$ is considered. filename:cold1.tex ON CALDERON'S REPRODUCING FORMULA A modification of Calderon's reproducing formula involving a finite Borel measure is considered. This modification gives rise to wavelet type representations of singular integral operators with constant coefficients on a real line and clarifies conditions under which a reproducing formula holds. Convergence of new representations in $L^p$-norm and almost everywhere is studied. An algorithm of approximating initial function by a Calderon-type construction with a perturbed data is suggested. To print out these files, please, get also: definiti.tex, mathchar.tex and vanilla.sty. The author would greatly appreciate any comments and information related to possible applications.
From: Kevin.Parrott@comlab.ox.ac.uk Subject: Wavelets and time series I am using a discrete wavelet transform to process two types (probably correlated) of time series from medical measurements and would appreciate advice/references/background material on how to extend conventional auto-spectrum and cross-spectrum density functions to make use of the DWT. Many thanks. Kevin Parrott Oxford University Computing Laboratory Wolfson Building, Parks Road, Oxford, OX1 3QD, UK. Tel 044-865-273889, email kevin.parrott@comlab.ox.ac.uk
From: xxia@afit.af.mil (Xianggen Xia) Subject: Preprint: (p)reprints from X. Xia and co-workers available 0. Preprint: Vector-Valued Wavelets and Vector Filter Banks Authors: Xiang-Gen Xia and Bruce W. Suter Dept. of Electrical and Computer Engineering Air Force Institute of Technology Wright-Patterson AFB, OH 45433-7765 E-mail: xxia@afit.af.mil Phone: (513)255-6565 ext. 4413 FAX: (513)476-4055 Abstract: In this paper, we introduce vector-valued multiresolution analysis and vector-valued wavelets for vector-valued signal spaces. We construct vector-valued wavelets by using paraunitary vector filter bank theory. In particular, we construct vector-valued Meyer wavelets that are band-limited. We classify and construct vector-valued wavelets with sampling property. As an application of vector-valued wavelets, multiwavelets can be constructed from vector-valued wavelets. We show that certain linear combinations of known scalar-valued wavelets may yield multiwavelets. The difference between vector-valued wavelets and multiwavelets is that vector-valued wavelets can not only decorrelate vector-valued signals in the time domain but also in the spatial-domain. We also present discrete vector-valued wavelet transforms for discrete vector-valued (or blocked) signals, which can be thought of as a family of unitary vector transforms. 1. Reprint: On Sampling Theorem, Wavelets, Wavelet Transforms, IEEE Trans. on Signal Processing, vol. 41, pp. 3524-3535, Dec. 1993. Authors: Xiang-Gen Xia and Zhen Zhang Abstract: In this paper, we classify all orthogonal interpolating wavelets (OIW) and prove that OIW with compact support have and only have one possibility that is the Haar wavelet. We also present a family of OIW that have exponential decay. With these OIW, an application is the computation of wavelet series coefficients of a signal by the Mallat algorithm. For signals that are not in multiresolution spaces, we estimate the aliasing error in the sampling theorem by using uniform samples. Remark: In the paper, the OIW was called cardinal orthogonal scaling functions (COSF). 2. Reprint: The Backus-Gilbert Method for Signals in Reproducing Kernel Hilbert Spaces and Wavelet Subspaces, Inverse Problems, vol. 10, pp. 785-804, 1994. Authors: Xiang-Gen Xia and M. Z. Nashed Abstract: The Backus-Gilbert (BG) method is an inversion procedure for a moment problem when moments of a function and related kernel functions are known. In this paper, we consider the BG method when, in addition, the signal to be recovered is known a priori to be in certain reproducing kernel Hilbert spaces (RKHS), such as wavelet subspaces. We show that better performance may be achieved over the original BG method. In particular, under the D-criterion the BG method with RKHS information for a sampled signal in wavelet subspaces can be completely recover the original signal, while the one without any additional information can only provide a constant-valued signal. 3. Preprint: Signal Extrapolation in Wavelet Subspaces, to appear in SIAM J. on Scientific Computation, Jan. 1995. Authors: Xiang-Gen Xia, C.-C. Jay Kuo and Zhen Zhang Abstract: The Papoulis-Gerchberg (PG) algorithm is well-known for band-limited signal extrapolation. We consider the generalization of the PG algorithm to signals in the wavelet subspaces in this research. The uniqueness of the extrapolation for continuous-time signals is examined, and sufficient conditions on signals and wavelet bases for the generalized PG (GPG) algorithm to converge are given. We prove that all Meyer wavelets satisfy the conditions, where a family of interpolating Meyer wavelets are also given. We also propose a discrete GPG algorithm for discrete-time signal extrapolation, and investigate its convergence. Numerical examples are given to illustrate the performance of the discrete GPG algorithm. 4. Preprint: On Wavelet Coefficient Computation with Optimal Prefiltering, to appear in IEEE Trans. on Signal Processing, Aug. 1994. Authors: Xiang-Gen Xia, C.-C. Jay Kuo and Zhen Zhang Abstract: Several issues on signal sampling and wavelet coefficient computation for a continuous time signal with orthogonal or biorthogonal wavelet bases are studied in this research. Discrete wavelet transform (DWT) is often used to approximate wavelet series transform (WST) and continuous wavelet transform (CWT), since they can be computed numerically. We first study the accuracy of the computed DWT coefficients obtained from the Mallat and Shensa algorithms as an approximation of the WST coefficients. Based on the accuracy analysis, we show a procedure to design optimal FIR prefilters used in the Shensa algorithm to reduce the approximation error. Then, by extending the sampling theorm in wavelet spaces, we identify a class of signals whose exact WST coefficients can be obtained by using the Shensa algorithm. We also derive formulas characterizing the aliasing error resulted from general signals not in the class. Finally, numerical examples are presented to show the performance of the optimal FIR prefilters. 5. Preprint: On Orthogonal Wavelets with Oversampling Property, to appear in J. of Fourier Analysis and Appl. Authors: Xiang-Gen Xia Abstract: In this paper, we consider orthogonal wavelets with oversampling property. We prove that if an orthogonal scaling function with exponential decay has the oversampling property then it has the sampling property, i.e., it takes values 1 at 0 and o at other integers, and therefore, an orthogonal scaling function with compact support has the oversampling property if and only if it is the Haar function. 6. Preprint: On the Limit of Sampled Signal Extrapolation with a Wavelet Signal Model Authors: Xiang-Gen Xia, Li-Chien Lin and C.-C. Jay Kuo Abstract: The extrapolation of sampled signals from a given interval using a wavelet model with various sampling rates is examined in this paper. We present sufficient conditions on signals and wavelet bases so that the discrete-time extrapolated signal converges to its continuous-time counterpart when the sampling rate goes to infinity. Thus, this work provides a practical procedure to implement continuous-time signal extrapolation, which is important in wideband radar and sonar signal processing, with a discrete one via carefully choosing the sampling rate and wavelet basis. A numerical example is given to illustarte the performance.
From: Fabrice Gras <fabrice@orcrist.fsc.qut.edu.au> Subject: Question: applications of wavelets to spectrum analysis Hello, I am looking for information on applications and theory of wavelet transform to spectrum analysis of nonstationary random processes. I would appreciate if someone working in this area could help me to find more information and references. Thank you. Fabrice Gras (email : f.gras@QUT.edu.au)
Note from the editor: In a discussion session at the SPIE meeting, held last week in San Diego, it was suggested to make quantization tables for wavelet based compression publicly available. This would facilitate comparisons between different schemes. The following message is a first reaction to this call. If you want to make your quantization table public, please send it to the Wavelet Digest and we would be happy to announce its existence and make it available through our information servers. --- From: jfs@SYBIL.risc.rockwell.com Subject: General: JPEG type quantization matrices for Wavelet Transforms Hey Wavletters: As promised at the San Diego SPIE conference, I have a reference to an article where JPEG type of coding was done for images using wavelet based methods (in this case for Wavelet Packets). This article is titled "JPEG-based Image Compression Scheme Using Fully-decomposed Wavelet Packet (FWP) Transform" by Kwo-Jyr Wong, Po-Yuen Cheng and C.-C. Jay Kuo found on pp. 75-86 of SPIE proceedings vol 2186 - Image and Video Compression (1994). They give a proposed JPEG type quantization matrix in the article, which is very interesting and informative. Jim Scholl jfs@risc.rockwell.com Rockwell Science Center Thousand Oaks, CA 91360
From: Praveen Kumar <praveen@hydromet.gsfc.nasa.gov> Subject: Question: Contributions to library of algorithms Looking for contributions to LIBRARY OF ALGORITHMS FOR WAVELET APPLICATIONS Chrysalis Software is a startup software company and the objective of one of the products being developed is to make a library of algorithms available for various aspects of wavelet analysis and applications in various fields such as mathematics, signal and image processing, geophysics, turbulence, medical signal processing and imaging etc. People during their course of work/research develop several algorithms and programs/routines which can be of use to other people in their research. Often, these programs are not large enough to be advertized or marketable and are consequently not put together in a user friendly form. However, several of these programs can be put together in a library and made available to people who may use these programs with little effort or include it as part of their own programming effort. Chrysalis Software is working on creating such a library. If you have routines/programs which you think are important/unique in some way that other people could use please get in touch with us directly by sending an e-mail to praveen@mykonos.safhl.umn.edu for more details. We will put such a library together and make it available at a "small price" and inturn pay royalties to the contributor(s) based on his/her contribution and number of copies sold. We are looking for response from two groups of people: 1) Those who are interested in contributing their programs to this library; and 2) Those who think this is a good idea and would be interested in buying such a library. This response will be used as a market survey only, to assess the market potential of this endeavor.
From: Praveen Kumar <praveen@hydromet.gsfc.nasa.gov> Subject: Book: Wavelets in geophysics Announcing "WAVELETS IN GEOPHYSICS" by Academic Press Inc. I would like to announce the availability of "Wavelets in Geophysics" , edited by Efi Foufoula-Georgiou and Praveen Kumar. It is volume 4 in Charles Chui's series "Wavelet Analysis and Its Applications". It can be ordered by calling 1-800-321-5068 Mon-Fri 8:30am to 7:00 pm Eastern time (Fax: 1-800-336-7377) with ISBN # 0-12-262850-0 priced tentatively at $59.95 (total pages 365). The following is a brief table of contents: 1) Wavelet analysis in geophysics: An introduction by Praveen Kumar and Efi Foufoula-Georgiou 2) Applications of structure preserving wavelet decompositions to intermittent turbulence: A case study by C. R. Hagelberg and N. K. K. Gamage 3) Intermittency in atmospheric surface layer turbulence: the orthonormal wavelet representation by G. G. Katul, J. D. Albertson, C. R. Chu, and M. B. Parlange 4) An adaptive decomposition: application to turbulence by J. F. Howell and L. Mahrt 5) Wavelet analysis of diurnal and nocturnal turbulence abaove a maize crop by Y. Brunet and S. Collineau 6) Wavelet spectrum analysis and ocean wind waves by P. C. Liu 7) Wavelet analysis of seafloor bathymetry by S. A. Little 8) Analysis of high resolution marine seismic data using the wavelet transform by C. J. Pike 9) Including multiscale information in the characterization of hydraulic conductivity distributions by K. E. Brewer and S. W. Wheatcraft 10) Wavelet based multifractal analysis of non-stationary and/or intermittent geophysical signals by A. Davis, A. MArshak and W. Wiscombe 11) Simultaneous niose suppression and signal compression using a library of orthonormal bases and the minimum description length criterion by N. Saito 12) Long memory processes, the Allan variance and wavelets by D. B. Percival and P. Guttorp
From: "Prof. E.B. Saff" <esaff@gauss.math.usf.edu> Subject: Contents: Constructive Approximation vol 10 nr 3 CONSTRUCTIVE APPROXIMATION Volume 10 Number 3 1994 301 Doron S. Lubinksy and Vilmos Totik Weighted Polynomial Approximation with Freud Weights 317 Clemens Markett Linearization of the Product of Symmetric Orthogonal Polynomials 339 S. Bloom and G. S. De Souza Weighted Lipschitz Spaces and Their Analytic Characterizations 377 D. P. Dryanov Polynomials of Minimal L_{\alpha}-Derivation, \alpha>0 411 P. Borwein and T. Erd\'{e}lyi Markov-Bernstein-Type Inequalities for Classes of Polynomials with Restricted Zeros 427 L. Frerick and J. M\"{u}ller Polynomial Approximation and Maximal Convergence on Faber Sets
From: "Bart M. ter Haar Romeny" <bart@cv.ruu.nl> Subject: Book: Geometry-Driven Diffusion in Computer Vision Recently there was an inquiry to the source of anisotropic diffusion. This field, where the process of evolution (blurring) of the image is a function of the image information itself, is rather new. A number of approaches to this process are currently developed. In September 1994 the book "Geometry-Driven Diffusion in Computer Vision" will appear, which is the first tutorial book in this field. The contributions are from the leading researchers in this area, in a transatlantic collaboration funded jointly by the NSF and Esprit. The book will be published by Kluwer. Information about availability can be acquired from Kluwer (Paul Roos, PhD, email: roos@wkap.nl). There will be a 25% discount for AMS members, and other societies is being considered. "GEOMETRY-DRIVEN DIFFUSION IN COMPUTER VISION" Editor: Bart M. ter Haar Romeny - Utrecht University, the Netherlands Contents: Foreword by Jan Koenderink - Utrecht University Introduction 1. Linear Scale-Space I: Basic Theory Tony Lindeberg - Bart M. ter Haar Romeny 2. Linear Scale-Space II: Early Vision Operations Tony Lindeberg - Bart M. ter Haar Romeny 3. Anisotropic Diffusion Pietro Perona - Taka Shiota - Jitendra Malik 4. Vector-Valued Diffusion Ross Whitaker - Guido Gerig 5. Bayesian Rationale for the Variational Formulation David Mumford 6. Variational Problems with a Free Discontinuity Set Antonio Leaci - Sergio Solimini 7. Minimization of Energy Functional with Curve-Represented Edges Niklas Nordstrom 9. Approximation, Computation and Distortion in the Variational Formulation Thomas J. Richardson - Sanjoy K. Mitter 9. Coupled Geometry-Driven Diffusion Equations for Low-Level Vision. Marc Proesmans, Eric Pauwels, Luc van Gool 10. Morphological Approach to Multiscale Analysis: From Principles to Equations Luis Alvarez - Jean Michel Morel 11. Differential Invariant Signatures and Flows in Computer Vision: A Symmetry Group Approach Peter J. Olver - Guillermo Sapiro - Allen Tannenbaum 12. On Optimal Control Methods in Computer Vision and Image Processing Benjamin Kimia - Allen Tannanbaum - Steven Zucker 13. Non-Linear Scale-Space Luc M. J. Florack, Alfons H. Salden, Bart M. ter Haar Romeny, Jan J. Koenderink, Max A. Viergever 14. A Differential Geometric Approach to Anisotropic Diffusion David Eberly 13. Numerical Analysis of Geometry-Driven Diffusion Equations Wiro J. Niessen - Bart M. ter Haar Romeny - Max Viergever Bibliography (370 references) Index (> 750 entries) Bart M. ter Haar Romeny, Editor Utrecht, the Netherlands April 1994

<webmaster@wavelet.org>