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

The Wavelet Digest - Volume 5, Issue 3 (April 29, 1996)


Today's Editor: Wim Sweldens
                wim@bell-labs.com


Today's Topics:

     1. Book:     The World According to Wavelets
     2. Preprint: Interpolating Subdivision for Meshes with Arbitrary Topology
     3. Preprint: A Pseudo-Wavelet Scheme for Navier-Stokes Equations
     4. Preprint: Boundary Conditions  of Wavelet-Galerkin 
     5. Preprint: Wavelet filtering with the Mellin transform
     6. Preprint: Speckle Noise Reduction with Symmetric Daubechies Wavelets
     7. Software: Xripple Wavelet analysis tool for one dimensional signals
     8. Software: MathWorks' Wavelet Toolbox
     9. Web:      Wavelet Cascade Applet (Java)
    10. Meeting:  Wavelets, operators and applications 
    11. Course:   Mathematics and Physics of Wavelets
    12. Contents: Special Issue on Wavelets of the Proceedings of the IEEE
    13. Answer:   Wavelets with binary coefficients (WD 5.2 #2)
    14. Answer:   Fast Continuous Wavelet Transform (WD 5.2 # 30)
    15. Answer:   Mistake in table in Chui's book? (WD 5.2 # 24)
    16. Question: Constant-Q filter in EEG signal analysis
    17. Question: Wavelets and differential equations
    18. Question: Wavelets and Communication Systems
    19. Question: Compactly supported biorthogonal box-spline wavelets
    20. Question: Wavelets and pattern recognition

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.

Back issues:
  ftp://ftp.math.sc.edu/pub/wavelet/archive
  http://www.math.sc.edu/~wavelet/

Current number of subscribers: 6235

Calendar of events:

 Apr 29-May 3: 1/f Long Range Dependent Processes, Les Houces, Fr.
*Jun 4-6:      Mathematics and Physics of Wavelets, College Park MD
 Jun 10-14:    Recent Advances in PDE's and applications, Venice, It
 Jun 18-21:    IEEE Symp. Time-Frequency/Time-Scale Analysis, Paris
 Jun 24-26:    Workshop on intelligent methods in signal proc., Spain
 Jun 24-27:    Internat. Congress Sound-Vibration, St. Petersburg
 Jun 26-28:    Time-frequency methods for economics & finance, Geneva
*Jul 24-28:    Wavelets: Operators and applications, UNC
 Aug 4-9:      SPIE, Wavelet applications, Denver
 Aug 4-9:      SPIE, Image compression, Denver
 Aug 31-Sep 4: Summer school on wavelets, Zakopane, Poland
 Sep 1-4:      IEEE DSP Workshop, Loen, Norway
 Sep 16-21:    Computational Modeling in Physics, Dubna, Russia
 Sep 4-6:      IEEE Workshop Neural Nets for Signal Proc., Kyoto
 Oct 9-11:     Second SIAM Conference on Sparse Matrices, Philadelph.
 Dec 17-19:    Mathematics in signal processing, Warwick, UK


From: akpeters@tiac.net (Klaus Peters) Subject: Book: The World According to Wavelets The World According to Wavelets: The Story of a Mathematical Technique in the Making by Barbara Burke Hubbard 1996, ISBN 1-56881-047-4, Hardcover, 264 pp., $34.00 available from A K Peters, Ltd., http://www.tiac.net/users/akpeters Written for a general audience this book tells the story of the meteoric rise of a new mathematical field from its scattered beginnings a decade ago to its current status as a hot topic cutting across traditional scientific boundaries. The material is masterfully organized so mathematical details can be assimilated at one's own pace; the main text is devoid of formulas and relates a story of people and ideas, while separate boxes and appendices contain intricate discussions for the more mathematically adventurous. The book just won the Prix d'Alembert, given by the Societe Mathematique de France. "This book exists on two levels: a first level of scientific information which is clear, lively, and well evoked, and a second level, much deeper, where the author tries to show us how scientists think and hesitate. There I find that she has created a very original work and one which may encourage a younger generation to embrace scientific voations. She has made the scientific life lively and accessible." Yves Meyer "Barbara Burke Hubbard's excellent book explains how a part of modern mathematics has contributed to modern technology, in words that show the pure delight of new discovery. ... The books organization welcomes readers regardless of mathematical preparation. The main section focuses on the history and some of the personalities that developed wavelet technology. Embedded in this are hypertext-like links to the appendices which explain the mathematical details of wavelets and why they have at once solved a diverse set of problems." M. V. Wickerhauser A K Peters, Ltd. 289 Linden Street Wellesley, MA 02181-5910 (617) 235-2210, Fax: (617) 235-2404 http://www.tiac.net/users/akpeters
From: Wim Sweldens <wim@bell-labs.com> Subject: Preprint: Interpolating Subdivision for Meshes with Arbitrary Topology Title: Interpolating Subdivision for Meshes with Arbitrary Topology Authors: Denis Zorin, Peter Schroeder, and Wim Sweldens Abstract: Subdivision is a powerful paradigm for the generation of surfaces of arbitrary topology. Given an initial triangular mesh the goal is to produce a smooth and visually pleasing surface, whose shape is controlled by the initial mesh. Of particular interest are interpolating schemes since they match the original data exactly, and are crucial for fast multiresolution and wavelet techniques. Dyn, Gregory, and Levin introduced the Butterfly scheme, which is known to yield C1 surfaces in the topologically regular setting. Unfortunately it exhibits various degeneracies in case of an irregular topology, leading to undesirable artifacts. We examine these failures and derive an improved scheme, which retains the simplicity of the Butterfly scheme, is interpolating, and results in smoother surfaces. Status: Technical Report CS-TR-96-06, Computer Graphics Group, Department of Computer Science, California Institute of Technology, 1996 To appear in Computer Graphics Proceedings (SIGGRAPH 96). The postscript file & color images can be retrieved from: http://www.cs.caltech.edu/~ps/zorin/
From: Philippe Charton <Philippe.Charton@univ-reunion.fr> Subject: Preprint: A Pseudo-Wavelet Scheme for Navier-Stokes Equations Preprint: A Pseudo-Wavelet Scheme for the Two-Dimensional Navier-Stokes Equations Authors: Philippe Charton and Valerie Perrier charton@univ-reunion.fr perrier@lmd.ens.fr Available at ftp://ftp.lmd.ens.fr/MFGA/pub/wavelets/ns.ps.Z Abstract: Numerical schemes taking advantage of the wavelet bases capabilities to compress both functions and operators are presented. Galerkin methods for the heat and Poisson equations in dimension two are developed and numerically tested. Then a pseudo-wavelet scheme, using collocation for the nonlinear term, is defined for the two-dimensional Navier-Stokes equations. Numerical tests prove its accuracy. Submitted to Matematica Aplicada e Computacional Philippe CHARTON Valerie PERRIER IREMIA LMD Universite de La Reunion ENS BP 7151 24, rue Lhomond 15, av. Rene Cassin 75231 Paris Cedex 05 97715 SAINT-DENIS MESSAG Cedex 9 France France
From: lu@ipc.akita-u.ac.jp (LU Dian Feng) Subject: Preprint: Boundary Conditions of Wavelet-Galerkin Preprint: Treatment of Boundary Conditions in the Application of Wavelet-Galerkin Method to a SH Wave Problem Author : Dianfeng LU, Tadashi OHYOSHI, Lin ZHU Abstract: One of main problems with the Wavelet-Galerkin is the treatment of boundary conditions. It is sloved by our method which is referred to as Fictitious Boundary Approach. In this approach, a fictitious boundary is assumed to tackle the difficulty of treating boundary conditions. And an additional condition is also made in order to ensure the real solution. This method is adapted to all the three kinds of boundary conditions in the theory of differential equation. To support our method, a SH wave mode problem in a plate is solved, and its numeral result is compared with the exact solution. Although the Wavelet-Galerkin has very good features for solving differential equations, we find that it is still difficult to represent the infinite resonance solution after investigating the error convergence near the resonance point. And we emphasize the importance of selecting a proper sacle $j$ especially to represent the solution near the resonance. This paper will be appeard in International Journal of The Society of Materials Engineering for Resources. The postscript file of this paper is available. Anyone can require it by mail. Comments would be very appreciated. Dianfeng LU Ohyoshi Laboratory Department of Mechanical Engineering Akita University E-Mail: lu@ipc.akita-u.ac.jp Tel : +81-188-89-2345 (Lab.) FAX : +81-188-36-3829 (home)
From: Gerald Kaiser <gkaiser@cs.uml.edu> Subject: Preprint: Wavelet filtering with the Mellin transform Preprint: Wavelet filtering with the Mellin transform Author: Gerald Kaiser, University of Massachusetts at Lowell To appear in Applied Mathematics Letters ABSTRACT: It is shown that any time-domain convolution operator can be represented exactly as a multiplication operator in the wavelet (time-scale) domain. The Mellin transform establishes a one-to-one correspondence between FREQUENCY FILTERS (multipliers in frequency) and SCALE FILTERS (multipliers in scale), subject to an ADMISSIBILITY CONDITION requiring the convergence of the defining integrals. This admissibility condition generalizes the usual one for the basic wavelet, reducing to the latter when the convolution operator being represented is the identity. Applications to the analysis and denoising of random signals are proposed. It is argued that the present method may be particularly useful for dealing with the effects of atmospheric turbulence because it is ideally suited for resolving spectral power laws.
From: lgagnon@mtl.unisysgsg.com (Langis Gagnon) Subject: Preprint: Speckle Noise Reduction with Symmetric Daubechies Wavelets Hi everyone, For those of you that are interested in noise reduction using wavelets, here is a new paper available upon request to Langis Gagnon (see email address below). TITLE: Speckle Noise Reduction of Airborne SAR Images with Symmetric Daubechies Wavelets AUTHORS: Langis Gagnon, R&D Department, LORAL Canada Inc., 6111 Ave. Royalmount, Montreal (Quebec), H4P 1K6, CANADA Fatima Drissi Smaili, Departement de Mathematiques, Universite de Montreal, C.P. 6128 Succ. Centre-Ville, Montreal (Quebec), H3C 3J7, CANADA SUMMARY: We report the study of a multiresolution speckle reduction method for airborne Synthetic Aperture Radar (SAR) images. The SAR image is subband-coded using Symmetric Daubechies (SD) wavelets followed by a noise estimate on the three high-pass bands. An elliptic wavelet coefficient thresholding rule (based on a principal component analysis) is applied, that preserves the global orientation of the complex wavelet coefficient distributions. Tests are performed on commercial airborne SAR images of uniform regions as well as on small dim objects. A speckle index is computed to quantify the noise reduction performance. We compare our results with those obtained using median and geometrical (Crimmins) filters. PUBLICATION: SPIE Proceedings # 2759, "Signal and Data Processing of Small Targets"Orlando, 1996 Langis Gagnon phone: (514)-340-8715 R&D Department FAX: (514)-340-8318 LORAL Canada Inc. email: lgagnon@mtl.unisysgsg.com 6111, Ave. Royalmount WEB: http://www.crm.umontreal.ca/~lgagnon Montreal (Quebec) H4P 1K6 CANADA
From: "K. Sunil Kumar" <nil@ee.iitb.ernet.in> Subject: Software: Xripple Wavelet analysis tool for one dimensional signals Xripple : Wavelet analysis tool for one dimensional signals Signal Processing and Artificial Neural Networks Laboratory Department of Electrical Engineering Indian Institute of Technology - Bombay Powai, BOMBAY 400 076 Contact: ubdesai@ee.iitb.ernet.in 091-22-5792545 x 2414 Xripple is a graphical user interface (GUI) for the wavelet analysis of one dimensional signals. This GUI has the following facilities. * Displays the discrete time wavelet transform and the symmetric wavelet transform along with the reconstructed signal of any given signal. * Displays scaling function as well as the mother wavelet. In addition it displays the magnitude response of the Fourier transform of the scaling function and the mother wavelet. This is helpful in getting to understand the time-frequency localization property. * Options to enable design of a "time dependent filter" by clicking the mouse. * Options to select the wavelet coefficients for reconstruction depending on their absolute value. * Includes features to store the transform coefficients, the scaling function, the mother wavelet and the selected transform coefficients for reconstruction and the reconstructed signal. * For starters some sample filter coefficients have been included with the database of the package (eg. Daubhechies, Coiflet). Option available to give the filter coefficients from an external file. * Online help. * Xripple has provision to incorporate any other module for time- frequency analysis of signals. At present the best basis wave packet program has some minor bugs ;-( we plan to rectify this in the near future. This GUI is developed using Qt libraries for GUI (free Linux version Qt-0.93). You must have the Qt GUI libraries with you to compile Xripple. You can ftp the Qt from the following site sunsite.unc.edu in /pub/Linux/devel/lang/c++/qt-0.93.tar.gz. The documentation for Qt can be browsed on the web at http://www.troll.no/ where you will also find other Qt-related links. How to get a copy of the Xripple? ftp 144.16.100.24 user: anonymous passwd: your_email_address cd pub hash bin get Xripple.tar.gz bye Ur_Machine> zcat Xripple.tar.gz |tar xvf - will have all the files necessay at your machine. You can if you are using linux directly start using the binary Xripple, else you will have to read the file called INSTALL which comes with the distribution and get the qt libraries. If you have any difficulty in downloading the file, please feel free to get back to us at this email addresses ubdesai@ee.iitb.ernet.in posted by sunil (nil@ee.iitb.ernet.in) for Professor U. B. Desai.
From: "Roy Lurie" <roy_lurie@mathworks.com> Subject: Software: MathWorks' Wavelet Toolbox Announcing The MathWorks' Wavelet Toolbox for use with MATLAB The MATLAB Wavelet Toolbox, from The MathWorks, provides a comprehensive collection of M-files and GUI based tools for examining local, multiscale, and nonstationary phenomena. Wavelet methods offer additional insight and performance in any application where Fourier techniques have been used. The toolbox is useful in many signal and image processing applications, including speech and audio processing, communications, geophysics, finance, and medicine. The Wavelet Toolbox is written by Michel Misiti, Yves Misiti, Georges Oppenheim and Jean-Michel Poggi, of the "Laboratoire de MathŽmatiques" Orsay-Paris 11 University, France. Bundled with the Wavelet Toolbox is the new book, "Wavelets and Filter Banks", by Gilbert Strang and Truong Nguyen. Many exercises in this book are drawn from the toolbox. Features : - Complete GUI and command line functionality for analysis, synthesis, de-noising, and compression of signals and images - Extensive user's guide introduces wavelet concepts and reinforces them with numerous examples and a complete reference section - Continuous wavelet transform for multiscale signal analysis - Analysis and synthesis of 1-D and 2-D discrete wavelet transform (DWT) - Multiresolution decomposition and analysis of signals and images - Wide selection of wavelet basis functions (user extensible) - 1-D and 2-D wavelet packet transforms - Entropy-based wavelet packet tree pruning for "best-tree" and "best-level" analysis - De-noising with soft and hard thresholding For more information: WWW : http://www.mathworks.com/wavelet.html E-mail: info@mathworks.com Phone : (508) 647-7000 Fax : (508) 647-7101
From: Wim Sweldens <wim@bell-labs.com> Subject: Web: Wavelet Cascade Applet (Java) You might have heard about Java, the new object-oriented computer language for Web programming. Java allows you to write ``Applets'' which are small program that run inside Web pages. I have written a Java applet which allows you to interactively graph all possible wavelets and scaling functions with four terms in the refinement relation. By simply moving your mouse inside the parameter space you can see the shape of the wavelet smoothly change from Haar to Linear to spline to D4. All graphs are computed with the cascade (or subdivision) algorithm. Operating instructions and some mathematical background are provided. In order to run the applet, you need version 2.0 or higher of Netscape. Then simply point your browser to /math/people/wim/cascade/ Note that it might take some time to download the Java classes. Please feel free to contact me with comments, suggestions, or bug-reports. I only ran the program on my machine (SGI), but one of the nice feature of Java is that it runs on all architectures. If this is not the case, please let me know. Also, I would like to encourage other people working on wavelet related applets to send them in to the Wavelet Digest. I will build a Web page with the pointers to the different submissions. Wim Sweldens ps: The Netscape navigator can be downloaded from http://www.netscape.com/ More information about Java can be found at http://java.sun.com/ pps: The rendering the mathematical symbols inside the Web page on the mathematical background is an other example of an applet.
From: David Royster <droyster@uncc.edu> Subject: Meeting: Wavelets, operators and applications First Announcement of the Conference on ``Wavelets, relations with operators and applications'' Dear Colleagues: We are organizing a conference on ``Wavelets, relations with operators and applications'' at UNC-Charlotte, July 24-28, 1996. The conference is sponsored by AFOSR, NSA(pending), NSF and UNC Charlotte. The conference will focus on wavelet theory and the mathematics behind the theory. Applications will occupy a considerable part of the conference. Speakers include Professors Charles Chui, David Larson and Guido Weiss. During the conference, we expect people will exchange questions across pure and applied mathematics. We sincerely hope that people will start new research partnerships. We plan to publish a proceedings. If you are interested to attend the conference, please send an e-mail message to xdai2@uncc.edu including your name, mailing address and phone number. Also, we have a limited number of slots for 30-minute talks. Should you be interested in giving a 30 minute talk, please send an abstract, your mailing address and your phone number to: Xingde Dai at xdai2@uncc.edu or snail mail to Xingde Dai Department of Mathematics UNC Charlotte Charlotte NC 28223-9998 or call (704) 547-2858. Priority will be given on a first come first served basis. The registration fee is $200 for commercial participants, $50 for non- commercial participants. Graduate students will receive a special rate of $20. Speakers will pay the registration fee. (We have no coffee and cookie money from the funding agencies). We have a block of rooms at Hampton Inn and Holiday Inn (at University) with a special UNC Charlotte rate. We have a modest budget on which to run this conference. Some funds come from AFOSR, NSA(pending), NSF and UNC Charlotte. We will make every effort to help everyone as much as possible, especially graduate students. The attendees are encouraged to find support from their institutions, grants or other sources. The priority for consideration for financial support will follow the reversal of seniority. Keep in touch with the conference on the conference WWW page: http://www.coe.uncc.edu/cas/wavelet_conf.html See you in Charlotte! Sincerely Yours, Xingde Dai Department of Mathematics UNC-Charlotte, Charlotte, NC 28223 (Office) (704) 547-2858 (Fax) (704) 510-6415 xdai2@uncc.edu xdai@uncc.edu
From: Gerald Kaiser <gkaiser@cs.uml.edu> Subject: Course: Mathematics and Physics of Wavelets Offered by The Applied Technology Institute Phone (410) 531-6034 fax (410) 531-1013 Email atiinfo@aol.com COURSE TITLE: Introduction to the Mathematics and Physics of Wavelets, with Applications to Remote Sensing JUNE 4-6, 1996 Washington DC Area Holiday Inn College Park Maryland INSTRUCTOR Professor Gerald Kaiser Mathematical Sciences Department, University of Massachusetts at Lowell TEXTBOOK: Participants will receive a copy of the instructor's popular book "A Friendly Guide to Wavelets" (Birkhauser, 1994). DESCRIPTION: 12 lectures, each 50 minutes, followed by 10 minutes for questions. The instructor will also be available for extended discussions. LECTURES: 1. Continuous time-frequency analysis 2. Continuous time-scale (wavelet) analysis 3. Generalized frames: Keys to analysis and synthesis 4. Time-frequency sampling theorems (Discrete time-frequency analysis) 5. Time-scale sampling theorems (Discrete time-scale analysis) 6. Multiresolution analysis 7. Subband filtering 8. FIR filters and Daubechies wavelet bases 9. Applications to atmospheric turbulence data analysis 10. Radar and sonar ambiguity functions: Canonical applications of time-frequency and time-scale analysis 11. Electromagnetic and acoustic wavelets: Space-time adaptive analysis for sensing and communication 12. Some new applications to remote sensing A second related course Modern Spectral Analysis will be presented by Dr Stephen Kay May 29-31 in the Washington DC Area. Call 410-531-6034 for more information. To register or to receive a full course catalog of all of ATI's courses phone the Applied Technology Institute at (410) 531-6034 or Fax (410) 531-1013. Enrollment is limited. For more information about the course content, contact "gkaiser@cs.uml.edu"
From: jelena@bell-labs.com (Jelena Kovacevic) Subject: Contents: Special Issue on Wavelets of the Proceedings of the IEEE A special issue on wavelets of the Proceedings of the IEEE appeared in April. It was co-edited by Jelena Kovacevic from Bell Labs and Ingrid Daubechies from Princeton University. It spans a wide range of topics and is meant to give an overview of the use of wavelets in various areas. Proceedings of the IEEE, Special Issue on Wavelets April 1996, Vol 84, No. 4 TABLE OF CONTENTS: Scanning the Special Issue on Wavelets J. Kovacevic and I. Daubechies Where Do Wavelets Come From? - A Personal Point of View I. Daubechies Wavelets: The Mathematical Background A. Cohen and J. Kovacevic Wavelets and Time-Frequency Analysis N. Hess-Nielsen and M.V. Wickerhauser Wavelets, Subband Coding and Best Bases K. Ramchandran, M. Vetterli and C. Herley Characterization of Acoustic Signals Through Continuous Linear Time-Frequency Representations P. Guillemain and R. Kronland-Martinet Emerging Applications of Multirate Signal Processing and Wavelets in Digital Communications G.W. Wornell Wavelets for a Vision S. Mallat Wavelets in Computer Graphics P. Schroeder A Review of Wavelets in Biomedical Applications M. Unser and A. Aldroubi Wavelets and Turbulence M. Farge, N. Kevlahan, V. Perrier and E. Goirand Wavelets and the Study of the Distant Universe A. Bijaoui, E. Slezak, F. Rue and E. Lega Wavelets: What Next? W. Sweldens
From: ali akansu fac ee <ali@hertz.njit.edu> Subject: Answer: Wavelets with binary coefficients (WD 5.2 #2) Ref: Some information on multilplierless subband/wavelet transforms (related to Prof. Gil Strang's announcement in WD 5.2 #2) The following information might be helpful to remind the waveletters some earlier work on multiplierless (power of two, etc) subband/wavelet transforms (orthogonal or bi-orthogonal) in the literature. 1) Multiplierless PR-QMFs were first presented in SPIE Visual Comm. and Image Processing, 1992(Boston), (A.N. Akansu, Multiplierless Suboptimal PR-QMF Design, Proc. SPIE Visual Comm. andImage Processing, 1992, Boston, pp. 723-734.) These filters were also mentioned in the book A.N. Akansu and R.A. Haddad, Multiresolution Signal Decomposition: Transforms, Subbands and Wavelets. Academic Press, 1992. More recently, there was a US patent issued as A.N. Akansu, Multiplierless 2-Band Perfect Reconstruction Quadrature Mirror Filter (PR-QMF) Banks, US Patent# 5,420,891, May 30, 1995. These filters and their extensions have been already used by some companies for image/video compression. 2) D. LeGall and Tabatabai have designed integer coefficient, short tap filters for image/video coding applications sometimes in late 1980's. (Akansu&Haddad book also revisits LeGall-Tabatabai filters) Their filters fit better to the bi-orthogonal subband/wavelet transforms. 3) On biorthogonal filter banks, I think, Prof. M. Kunt's group at EPFL, Switzerland also reported some efficient solutions(I don't remember exactly where it was presented). 4) Most likely, I am missing some other work reported earlier in subband transform literature which perfectly fit under the new wavelet terminology around. A piece of information which I thought might be enlightening some waveletters. Regards, Ali N. Akansu ali@hertz.njit.edu
From: goswami@decwa.ece.uiuc.edu (Jaideva C. Goswami) Subject: Answer: Fast Continuous Wavelet Transform (WD 5.2 #30) This is in response to the question regarding the existence of fast continuous wavelet transform algorithm (WD 5.2 #30) . The following three papers summarizes our work on such algorithm: (1) Charles K. Chui, Jaideva C. Goswami, and Andrew K. Chan, ``Fast integral wavelet transform on a dense set of time-scale domain,'' Numer. Math., Vol.70, No.3, pp.283-302, 1995. (2) J. C. Goswami, A. K. Chan, and C. K. Chui, ``An application of fast integral wavelet transform to waveguide mode identification,'' IEEE Transactions on Microwave Theory and Techniques, pp.655-683, March 1995. (3) J. C. Goswami, A. K. Chan, and C. K. Chui, ``On a spline-based fast integral wavelet transform algorithm,'' in Ultra-wideband short pulse electromagnetics 2, L. Carin and L. B. Felsen, Eds., New York: Plenum Press, pp. 455-463, 1995. Jaideva C. Goswami Electromagnetic Communication Laboratory Department of Electrical and Computer Engineering University of Illinois 1406 W. Green Street, Urbana, IL 61801-2991 Phone (217)-244-0178 (Work) (217)-344-8480 (Home) Fax (217)-333-8986 e-mail: goswami@decwa.ece.uiuc.edu
From: goswami@decwa.ece.uiuc.edu (Jaideva C. Goswami) Subject: Answer: Mistake in table in Chui's book? (WD 5.2 #24) This is in response to a question regarding accurarcy of values of the decomposition sequence for linear spline case in Chui's book (WD 5.2 #24). There is no mistake in the table. Here is a simple way to verify the accuracy (equation numbers mentioned below are from Chui's book). (1) Coefficients {a_k} and {g_k} are related (see, for example, eqn. 6.5.2). (2) Consider 6.2.1 (this equation is equivalent to 6.5.1). (3) Expression for E_phi and P(z) are known, e.g., for m = 2 (linear spline) 6z E_phi(z) = 1 + 4z + z^2 ; 4P(z) = 1 + 2z + z^2. (4) Write 1 / E_phi(z) = \sum_k r_k z^k. Expressions for r_k for a spline of order m can be seen in [1,2]. In particluar, for m = 2 r_k = r_{-k} = (-1)^k \sqrt(3) (2 - \sqrt(3))^k ; k = 0, 1, .... (5) The coefficients {g_k} and hence {a_k} can be found by comparing the powers of z on both sides of 6.2.1. (6) Finally, observe that \sum_k a_k = 1 (can be verified using 6.2.1 and the partition of unity property of spline). If we simply add all the coefficients in the table given in Chui's book, we obtain 1. While adding we must take care of the negative indices. More precisely, for the present case \sum_k a_k = a_1 + 2 \sum_{k=2}^{k=21} a_k = 1.0 From the above also it is clear that a_1 could not be 0.4330127 as mentioned in the question if the rest of the coefficients are correct. [1] Jaideva C. Goswami, Applications of Semi-Orthogonal Spline Wavelets in Electromagnetics and Microwave Problems, Ph.D. Dissertation, Texas A&M University, College Station, August 1995. [2] J. C. Goswami, A. K. Chan, and C. K. Chui, ``On a spline-based fast integral wavelet transform algorithm,'' in Ultra-wideband short pulse electromagnetics 2, L. Carin and L. B. Felsen, Eds., New York: Plenum Press, pp. 455-463, 1995. Jaideva C. Goswami Jaideva C. Goswami, Ph.D. Electromagnetic Communication Laboratory Department of Electrical and Computer Engineering University of Illinois 1406 W. Green Street, Urbana, IL 61801-2991 Phone (217)-244-0178 (Work) (217)-344-8480 (Home) Fax (217)-333-8986 e-mail: goswami@decwa.ece.uiuc.edu
From: Dirk Walterspacher <dwalters@bme.jhu.edu> Subject: Question: Constant-Q filter in EEG signal analysis I am a student working in the field of biomedical engineering and at the moment I am writing my thesis about EEG signal processing. I am working with energy operators to track the energy of the signal and to estimate the instanteanous amplitude and frequency. To use the algorithms, a proper filtering of the signal is necessary. As a constant-Q filter, or multiscale wavelet-like filter bank is proposed in several articels, my questions are: 1) Do you know how to create such a filter bank (or better) 2) Do you know if there is an already existing file in some matlab toolbox doing this 3) Has somebody used such a filter bank for EEG signals It would by a great help, if somebody could answer me these questions. Thank you. dwalters@bme.jhu.edu
From: Anthony Galli <galli@ecn.purdue.edu> Subject: Question: Wavelets and differential equations I have a question concering wavelets and differential equations. I am trying to apply wavelet techniques to the solution of systems of linear ODE's as part of my phd research. Is there anyone there in wavelet land who has had success with this? Are there some references that have escaped my observation? Any criticism would be appreciated. Thanks in advance. Regards, A. W. Galli galli@ecn.purdue.edu
From: sinhar@liman.Rutgers.EDU (Rajnish Sinha) Subject: Question: Wavelets and Communication Systems Hello, I've found very few papers on applications of Wavelets to communication theory. I would really appreciate it if somebody has more information to share regarding this topic, thank you, Rajnish Sinha Wireless Information Networks Laboratory (WINLAB) Rutgers University NJ (908) 878-0408 email: sinhar@winlab.rutgers.edu
From: Wim Sweldens <wim@bell-labs.com> Subject: Question: Compactly supported biorthogonal box-spline wavelets I am looking for work on biorthogonal compactly supported box-spline wavelets. More particularly I am interested in the case where both the wavelet and the dual wavelet are compactly supported, but the dual wavelet need not be a linear combination of box-splines. Basically this would be the generalization of the Cohen-Daubechies-Feauveau biorthogonal spline wavelets to the box-spline setting. Thank you. Wim Sweldens wim@bell-labs.com
From: Kevin Englehart <kengleha@unb.ca> Subject: Question: Wavelets and pattern recognition Dear Digest, I am interested in determining the reason for the apparent sparsity of wavelet research as applied to pattern recognition. Is there an intrinsic problem with wavelet-based coefficients (or features derived thereof) that preclude their use as a feature set for pattern recognition or classification applications? Is there any work that addresses this issue? Regards, Kevin Englehart Institute of Biomedical Engineering University of New Brunswick, Fredericton, N.B., CANADA kengleha@unb.ca [Answer]

<webmaster@wavelet.org>