[Home] [Back Issues] [Search] [Calendar] [Vol 1] [Vol 2] [Vol 3] [Vol 4] [Vol 5] [Vol 6]

The Wavelet Digest, Volume 6, Issue 8 (August 12, 1997)




Today's Editor: Wim Sweldens
                Lucent Technologies, Bell Laboratories
                wim@bell-labs.com



Today's Topics:

     1: Book:     Interpolation, Identification and Sampling
     2: Preprint: Compactly supported (bi)orthogonal wavelets...
     3: Preprint: Wavelet and Gabor frames: The whole series 
     4: Preprint: Analysis of a Physiological Time Series using Wavelets
     5: Award:    Spelman mathematician receives distinguished writing award
     6: Software: Time-frequency analysis
     7: Course:   Multirate Digital Filters and Applications
     8: Meeting:  Wavelet session at Geophysics (GAU) Fall meeting
     9: CFP:      Biomedical signal processing
    10: Job:      PhD fellowship at Sheffield Hallam University
    11: Job:      Postdoc position in Image Databases at NEC Research
    12: Contents: J. Approx. Th. - jul97 TOC
    13: Contents: Constructive Approximation, Volume 13, Number 4, 1997
    14: Contents: Advances in Computational Mathematics 7-4
    15: Answer:   Construction almost tight frames (WD 6.7 #23)
    16: Answer:   Question to the bio-medical world (WD 6.7 #19)
    17: Answer:   Wavelets in bio-world (WD 6.7 #19)
    18: Answer:   Question on M-band wavelets (WD 6.7 #20)
    19: Question: Non-separable wavelet reading 
    20: Question: Machine Duty Cycle Analysis using Wavelets
    21: Question: On-line implementation of Wavelet Filtering.
    22: Question: Concerning lifting (WD 6.5 #21)
    23: Question: Which wavelet in which application ?
    24: Question: Edge detection in noisy images
    25: Question: Shiftability
    26: Question: Wavelets in Speech Recognition
    27: Question: Complex wavelets
    28: Question: Pitch detection in music signal
    29: Question: Looking for a public domain software  
    30: Question: 3D wavelet software.
     
Current number of members: 9997

E-mail:
   Help:                        help@wavelet.org
   Add yourself as a member:    add@wavelet.org
   Remove yourself as a member: remove@wavelet.org
   Publish in the next WD:      publish@wavelet.org

Visit us on the Web at http://www.wavelet.org/

Calendar of events:

 Aug 11-15: Tutorial on Wavelets in Statistics, Switserland
 Aug 27-29: Time-Frequency and Time-Scale, Coventry, UK
 Sep 14-17: Time-Frequency Wavelet analysis symposium, Sacramento
 Sep 24-26: IEEE Neural Networks for Signal Processing, Florida
*Oct 8-10:  UCLA Course: Multirate Digital Filters, Los Angeles
 Oct 12-13: Wavelets in statistics, Durham, NC
 Nov 10-12: Course by Gil Strang in Braunschweig, Germany
 Nov 16-Dec 6: UCLA short courses on communications, LA
*Dec 8-12:  Wavelets at Geophysics (GAU) Meeting, San Francisco
 Dec 17-19: Introduction to wavelets, Boston
 Jan  3-6:  9th Approximation Conference, Nashville, TN








Date: Tue, 12 Aug 1997 10:11:47 +0100 (BST)
From: J R Partington <pmt6jrp@amsta.leeds.ac.uk>
Subject: #1 Book: Interpolation, Identification and Sampling

                            New Book

Interpolation, Identification and Sampling, by Jonathan R. Partington.

London Mathematical Society Monographs, no. 17, Oxford University
Press, July 1997. xii+267pp. 18 figures. ISBN 0-19-850024-6.

This book presents applications of functional analysis and operator
theory to problems of interpolation of analytic functions, systems
identification, and sampling of signals.

1. Function spaces and operators --- Fourier series and approximate
identities, Fourier and Laplace transforms, Function spaces on the
disc, Function spaces on the half-plane, Blaschke products and zero
sets, Toeplitz and Hankel operators.

2. Interpolation and recovery --- Classical interpolation problems,
Zero sets for the disc algebra, Extension and recovery in the disc
algebra, Totik's recovery theorem, Interpolating sequences.

3. Worst-case identification of functions --- Causal operators and
linear systems, Worst-case identification in the disc algebra,
Two-stage algorithms, Model sets and Chebyshev identification,
Band-limited identification on the circle.

4. Approximation and interpolation --- $n$-widths and approximation,
Hankel-norm approximation, Bases and decompositions in the disc
algebra, IBC and optimal algorithms, Interpolation and validation.

5. Hilbert space methods --- Reproducing kernel Hilbert spaces, The
matching pursuit algorithm, Minimal-norm interpolation, Least squares
methods, Frames and wavelets.

6. Convolution operators --- Worst-case identification of
convolutions, Probabilistic inequalities, Chebyshev identification and
input design, Identification in $\ell _2$, IBC and validation.

7. Sampling --- The Paley-Wiener spaces, Regular sampling, General
sampling, Balayage.

8. Applications --- Linear systems, $H_\infty$ control, Examples of
identification.

Contributed by Jonathan Partington (j.r.partington@leeds.ac.uk)



Date: Wed, 16 Jul 1997 18:21:18 +0800 (SST)
From: "Zuowei Shen" <matzuows@math.nus.sg>
Subject: #2 Preprint: Compactly supported (bi)orthogonal wavelets...

Title: Compactly supported (bi)orthogonal wavelets generated by
       interpolatory  refinable functions

Authors: Hui Ji and  Zuowei Shen
         Department of Mathematics
         National University of Singapore
         Singapore 119260

Abstract:

This paper provides several constructions of compactly supported
wavelets generated by interpolatory refinable functions.  It was shown
in \ref{D1} that there is no real compactly supported orthonormal
symmetric dyadic refinable function, except the trivial case; and also
shown in \ref{L} and \ref{GM} that there is no compactly supported
interpolatory orthonormal dyadic refinable function.  Hence, for the
dyadic dilation case, compactly supported wavelets generated by
interpolatory refinable functions have to be biorthogonal wavelets.
The key step to construct the biorthogonal wavelets is to construct a
compactly supported dual function for a given interpolatory refinable
function.  We provide two explicit iterative constructions of such
dual functions with desired regularity. When the dilation factors are
larger than $3$, we provide several examples of compactly supported
interpolatory orthonormal symmetric refinable functions from a general
method.  This leads to several examples of orthogonal symmetric
(anti-symmetric) wavelets generated by interpolatory refinable
functions.

To get the .ps.Z file:
    http://www.math.nus.sg/~matzuows/preprints.html
or by anonymous ftp  from approx.math.ualberta.ca in directory 
   pub/Shen as the file shenji1.ps.Z



Date: Mon, 21 Jul 1997 20:58:57 +0800 (SST)
From: "Amos Ron" <amos@singa.math.nus.sg>
Subject: #3 Preprint: Wavelet and Gabor frames: The whole series 

Fiberization of wavelet, Gabor, and other systems: the entire series

Zuowei Shen and I (Amos Ron) had just finished our fifth (and last?) paper
is this series.   All are available as .ps and .ps.Z from
ftp://ftp.cs.wisc.edu/Approx
file names are provided. There is also a read.me file at that location. 
So here is the list:

paper1:  file:    frame1.ps
         title:   Frames and stable bases for shift-invariant subspaces
                  of $L_2(R^d)$.
         journal: Canadian J. Math.  47 (1995)
         content: The theory of fiberization. Very technical. Read
                  if you intend to use fiberization yourself.

paper2:  file:    wh.ps
         title:   Weyl-Heisenberg frames and Riesz bases in $L_2(R^d)$.
         journal: Duke Math. Journal (1997)
         content: applications of fiberization to Gabor systems: the 
                  duality principle of such systems, gramian analysis
                  and zak transform analysis. do not read if your 
                  interest is squarely in wavelet systems.

paper3  file:     affine.ps
        title:    Affine systems: the analysis of the analysis operator
        journal:  J. Functional Analysis (1997)
        content:  the theory of affine frames: quasi-affine fiberization,
                  complete characterization, affine product, the
                  rectangular extension principle, compactly supported
                  spline tight frames. read only if you are interested
                  in wavelets.

paper4  file:     dframe.ps
        title:    Affine systems II: dual systems
        journal:  new (the paper, not the journal)
        content:  complements paper3 with  a study of the
                  system-dual system setup, for affine frames and
                  affine Riesz basis.

paper5  file:     tight.ps
        title:    Compactly supported tight affine spline frames
                  in $L_2(R^d)$.
        journal:  Math. Comp. (1997)
        content:  in the title. for example, we derive a tight frame
                  from the Powell-Zwart element (bivariate C^1
                  piecewise quadratic on a 4-direction mesh, whose
                  shifts are neither Riesz basis nor a frame). the
                  algorithm derives a compactly supported tight frame
                  from any refinable box spline.

bonus  file:      cg.ps, co-author: Charly Gr\"ochenig
       title:     Tight compactly supported wavelet frames of
                  arbitarily high smoothness.
       jounral:   Proceedings of  AMS (1997)
       content:   it extends the paper5 construction to arbitrary dilation
                  matrices. the wavelets, though, are not piecewise-polynomials.



Date: Thu, 31 Jul 1997 16:58:51 -0300 (ARST)
From: Alejandra Figliola <figliola@ulises.ic.fcen.uba.ar>
Subject: #4 Preprint: Analysis of a Physiological Time Series using Wavelets

Title: Analysis of a Physiological Time Series using Wavelet
       Transform.

Authors: Alejandra Figliola and Eduardo Serrano

Instituto de Calculo, Facultad de Ciencias
Exactas y Naturales.(UBA)\\Pab.I Ciudad Universitaria. 
(1428) Buenos Aires. Argentina.
e-mails: figliola@ulises.ic.fcen.uba.ar and eduser@mate.uba.ar}} 

Abstract

A new method based on the Wavelet Transform Theory, to quantify and
visualize the order of a system, is present. It is applied to
analyzing the data set of the '1991 Santa Fe Time Series Prediction
and Analysis Competition', that consists of 4 hours and 43 minutes of
simultaneous heart rate, variation chest volume, blood oxygen
saturation, and EEG sleep stage, of a patient with apneas,
i. e. periods in which he takes a few quick breaths and then he stops
breathing, during the sleeping period. The phenomenon is non
stationary, because it not conserve in time the value of its mean
value and variance, and the three series are contaminated with
noise. Apneas, intermittent apneas and regular behaviors alternate
during the process and the data values appear in a complex
structure. Wavelet techniques are very powerful to detect changes in
the signals. Also, since the wavelet coefficients it is possible to
define the Information Cost Function (ICF) which is capable to
evaluate the order (or disorder) of a dynamic al system, also when its
time series is non stationary and noised.

Published in: IEEE Engineering in Med. & Biol., 16, N 3, 
May/June 1997



Date-warning: Date header was inserted by POLLUX.DAVIDSON.EDU
From: Stephen Davis <stdavis@davidson.edu>
Subject: #5 Award: Spelman mathematician receives distinguished writing award

Press release:

SPELMAN MATHEMATICIAN RECEIVES DISTINGUISHED WRITING AWARD

Colm Mulcahy received one of the two the Carl B. Allendoerfer Awards
given on August 2, 1997, at the Summer Mathfest in Atlanta.
Established in 1976, the Carl B. Allendoerfer Awards, consisting of a
citation and cash prize, are presented by the Mathematical Association
of America for articles of expository excellence published in
Mathematics Magazine.

Prof. Colm Mulcahy, Spelman College, received this prestigious award
for his article, "Plotting and Scheming with Wavelets" (Mathematics
Magazine 69, December, 1996).  As noted in the citation ...

        Mulcahy states "Wavelets are acquiring a visibility and
        popularity that may soon be on the scale first enjoyed
        by fractals a few years back."  Mulcahy's article is an
        admirable piece of work, presenting this topic now at a
        level fully accessible to undergraduate students and to
        the uninitiated mathematicians of all backgrounds.
        Mulcahy writes with clarity, charm and humor so that the
        reader is irresistibly drawn into fundamental problems
        and solutions for both humans and computers.  He earned
        his B.Sc. and M.Sc. in 1978 and 1979, respectively, from
        University College Dublin in his native Ireland.  His
        doctoral work was done at Cornell University and at
        Dortmund Unversitat in germany under the guidance of
        Alex F.T.W. Rosenberg.

The Mathematical Association of America (MAA) is the world's largest
organization devoted to mathematics education at the collegiate level.
The nearly 30,000 members of the MAA participate in a variety of
activities that foster mathematics education, professional
development, student involvement, and public policy.  MAA's national
focus is complemented by its 29 regional sections---together
functioning as an extensive network for the mathematics community.

MAA contact:
   Mary-McLean Bancroft, 202-387-5200, mbancrof@maa.org
   The Mathematical Association of America
   1529 Eighteenth Street, NW
   Washington, DC  20036

Prof. Mulcahy can be reached at colm@spelman.edu and
http://www.spelman.edu/~jcf

Stephen Davis
  Secretary-Treasurer, MAA Southeastern Section
  Dept of Mathematics, Davidson College, Davidson, NC  28036-1719
  office/voice mail: 704-892-2343 * home: 704-892-1394
  e-mail: stdavis@davidson.edu * fax: 704-892-2005
  web: http://www.davidson.edu/academic/math/davis/index.html



Date: Wed, 6 Aug 97 08:51:41 UT
From: "Paul Stringer" <idsldev@msn.com>
Subject: #6 Software: Time-frequency analysis

I would like to announce new software which is available for
performing multiresolution time-frequency analysis on time-domain
signals using wavelets and wavelet packets. The software is called
WavePak and is available for PCs on Windows 3.1x and Windows 95
platforms.

The following functionality is provided:-
- 17 real (one-dimensional) orthogonal wavelets
- wrap and zero-pad process methods
- wavelet basis
- adaptive wavelet packet basis
- automatic or manual steering between time / frequency resolution
- up to 32K samples can be processed in one transform
- output is calibrated to a linear Hz scale
- scaling options (e.g. lin / log, mag / power)
- graphical representation of the time-frequency plane (colour or gray-scale)
- "wizard" mode for beginners

There is a tutorial on "How Wavelets Work" intended for wavelet
beginners which comes with the software. This tutorial covers basic
aspects of continuous and discrete wavelet theory and implementation,
and is designed to help users understand the software.

Signals can be imported by a wide range of methods (e.g. WAV files, MS
Excel) or can be generated by hardware. Other DSP and data
visualisation tools are also available.

For more info, contact:

Sales: Alan Benetts, <100666.351@compuserve.com>
Technical: Toby Sharp, <idsldev@msn.com>



Date: Mon, 14 Jul 1997 18:29:24 -0700
From: "Goodin, Bill" <bgoodin@unex.ucla.edu>
Subject: #7 Course: Multirate Digital Filters and Applications

 On October 8-10, 1997, UCLA Extension will present the short course,
 "Multirate Digital Filters and Applications", on the UCLA campus in
 Los Angeles.
 
 The instructor is Professor frederick harris, Electrical and Computer
 Engineering, San Diego State University.
 
 This course is an introduction to multirate digital filters, which
 are variants of non-recursive filters, and incorporate one or more
 resamplers in the signal path.  These embedded resamplers affect
 changes in sample rate for upsampling, downsampling, or combinations
 of both.  Changes in sampling rate as part of the signal processing
 is a feature unique to sampled data systems. and has no counterpart
 in continuous signal processing.  Benefits include reduced cost for a
 given signal processing task and improved levels of performance for a
 given computational burden.  This economy of computation has become
 an essential requirement of modern communication systems,
 particularly battery-operated equipment.
 
 Specific course topics include: Introduction to sample rate
 conversion, Non-recursive (finite impulse response) filters,
 Prototype FIR filter design methods, Decimation and interpolation,
 Multirate filters, Two-channel filter banks, M-channel filter banks,
 Proportional bandwidth filter banks and wavelet analysis, Polyphase
 recursive all-pass filter banks, Multirate filter applications.
 
 UCLA Extension has presented this highly successful short course 
 since 1995.
 
 The course fee is $1195, which includes extensive course materials.  
 These materials are for participants only, and are not for sale.
 
 For additional information and a complete course description, please
 contact Marcus Hennessy at:
 
 (310) 825-1047
 (310) 206-2815  fax
 mhenness@unex.ucla.edu
 http://www.unex.ucla.edu/shortcourses/
 
 This course may also be presented on-site at company locations.
 


Date: Tue, 29 Jul 1997 15:34:46 -0500 (CDT)
From: kumar praveen <kumar1@ux1.cso.uiuc.edu>
Subject: #8 Meeting: Wavelet session at Geophysics (GAU) Fall meeting

         American Geophysical Union Fall 1997 Meeting
        (December 8-12, 1997, San Francisco, California)

                    Special Union Session

       Advances in Wavelet Transform Applications in Geophysics

Our goal is to bring together scientists who use wavelet analysis for
the understanding, modeling, and description of geophysical processes.
Papers are solicited that present innovative use of multiscale
analysis/modeling techniques and demonstrate their benefits for
improved understanding and prediction of geophysical processes

Conveners:  

Praveen Kumar
Department of Civil Engineering, University of Illinois
205 North Mathews Avenue, Urbana, IL 61801 
(217-333-4688; kumar1@uiuc.edu)

Efi Foufoula-Georgiou
St. Anthony Falls Laboratory, University of Minnesota
Mississippi River and 3rd Ave. S.E., Minneapolis, MN 55414 
(612-627-4595; efi@mykonos.safhl.umn.edu).

Abstracts must be received by AGU headquarters no later than August
27, 1997 for postal/express mail and e-mail abstracts

or

September 3, 1997 by 11:59 P.M Eastern Time 
for Interactive Web Form abstracts see http://www.agu.org

See AGU web site, http://www.agu.org, or EOS, Transactions of American
Geophysical Union, Vol. 78, No. 26, July 1, 1997 for abstract format
and submission details

Send a copy of the abstract to one of the conveners irrespective of
whether you submit your abstract to AGU by e-mail, WWW or regular
mail.



Date: Wed, 30 Jul 97 19:14 BRT
From: "Luciano da F.Costa" <luciano@ifqsc.sc.usp.br>
Subject: #9 CFP: Biomedical signal processing

                             Call for papers

                        Applied Signal Processing

               Special Issue On Biomedical Signal Processing

Guest Editors:

Luciano da F. Costa, 
Cybernetic Vision Research Group,
DFI-IFSC, University of Sao Paulo
Caixa Postal 369,
Sao Carlos, SP
13560-970
Brazil
Tel: +55 162 74 9199
Fax: +55 162 71 3616
E-Mail: luciano@ifqsc.sc.usp.br

William Sandham,
Department of Electronic and Electrical Engineering,
University of Strathclyde
Glasgow G1 1XW
Scotland
Tel: +44 141 548 2406
Fax: +44 141 552 2487
E-Mail: w.sandham@eee.strath.ac.uk

The last two decades of scientific development have been characterised
by major advances in biomedical technology. Genetic and biomolecular
engineering have progressed from science fiction to reality, and
diagnostic techniques such as CT, MRI, PET and SPECT, have evolved
from basic research to widespread use in hospitals. Major
breakthroughs have been made in the neural sciences, which have led to
a better understanding of biological neural structures, and the
engineering of more powerful and versatile artificial neural networks
and models.

As we enter the 21st century, these technologies are destined to reach
a level of development which will redefine many concepts and alter the
human perspective in an unprecedented way.  Many of these developments
have been made possible through major advances in theoretical and
applied signal processing.  Deterministic methods for analysis and
synthesis, such as those based on the short-time Fourier transform and
wavelets, have now reached maturation, and excellent progress has been
made using statistical methods such as Markov random fields,
regularization theory, and Kalman filtering, for noise suppression,
signal enhancement and optimisation.

The advent of new biomedical devices, sensors and families of generic
and dedicated processors have also proven to be essential enabling
technologies for practical applications of such increasingly complex
techniques.

This Special Issue will review the current status of biomedical signal
processing, and address future techniques and applications.
Submissions should either provide a comprehensive perspective on
topics of special relevance (see list below) or report novel advances
in the theory and application of methods and architectures for
biomedical signal processing.

Papers are invited, but are not limited, to the topics listed below:

Biomedical Topics:
        * ECGs and EEGs
        * Event related (evoked) potentials
        * Orphometry
        * DNA analysis
        * Medical imaging
        * Tomography and optical imaging
        * Microscopy
        * Biomagnetism
        * Echocardiography
        * Radiology and Radiotherapy
        * Diagnostic ultrasound
        * Biophysics
        * Biosensors

Signal Processing Topics:
        * Medical informatics
        * Mathematical modeling and simulation
        * Measurement and medical instrumentation
        * Image processing, visualization and computer vision
        * Computational neuroscience
        * Artificial neural networks
        * Chaos and fractals
        * Wavelets and time-frequency analysis
        * Higher-order statistics
        * Non-linear analysis
        * Adaptive techniques

Paper Submission:
Prospective authors should prepare their manuscripts in the accepted 
ASP format, and send 4 copies to Luciano Costa at the above address. 
Papers will be reviewed by senior researchers from industry and academia.

Schedule:
Paper Submission Deadline               28 February 1998
Notification of Acceptance/Rejection    30 May 1998
Final Submission Deadline               27 August 1998



Date: Fri, 11 Jul 97 15:45:00 0BS
From: "Saatchi, Reza" <r.saatchi@shu.ac.uk>
Subject: #10 Job: PhD fellowship at Sheffield Hallam University

A PhD studentship jointly sponsored by Sheffield Hallam University and
Sheffield University is currently available. The work involves
development of novel signal processing methods including
multiresolution analysis for signal detection and
characterisation. The work will also include application of artificial
intelligence techniques for pattern recognition.

The post is for three years, commencing from September 1997. It is
suitable for a person with a degree with a 1st or a 2-I in
electronics, physics, mathematics or any other related area.

The salary is ?5300 per year. The tuition fees will be paid and there
is also an allowance for consumables. For more details please contact:

Dr R Saatchi
School of Engineering
Sheffield Hallam University
Pond Street
Sheffield
S1 1WB
UK.
Tel:       +44 (0)114 253 3265
E-Mail: R.SAATCHI@SHU.AC.UK



Date: Tue, 22 Jul 1997 14:02:31 -0400
From: hstone@research.nj.nec.com (Harold Stone)
Subject: #11 Job: Postdoc position in Image Databases at NEC Research

Position PostDoc Position in Image Databases at NEC Research Institute

Applications are invited for a two-year PostDoctoral Researcher
position in the area of image databases at NEC Research Institute. The
successful candidate will have a Ph. D. and experience in image
processing with knowledge of wavelet and Fourier transforms, image
compression, and a strong background in mathematics and statistics to
support research in image searching techniques.

The Institute, located in Princeton, New Jersey, has close ties with
many outstanding research universities located in the area.  The
Institute's laboratories are state-of-the-art and include several
high-end parallel computer servers.  For more details about the NEC
Research Institute, please see http://www.neci.nj.nec.com/.

Applicants must show documentation of eligibility for employment.  NEC
is an equal opportunity employer.  Interested applicants should send
their resumes, along with copies of selected publications and names of
three references, to Harold S. Stone, NEC Research Institute, 4
Independence Way, Princeton, NJ, 08540 or hstone@research.nj.nec.com.



Date: Sun, 20 Jul 1997 20:08:32 -0400 (EDT)
From: Edit Kurali <kurali@math.ohio-state.edu>
Subject: #12 Contents: J. Approx. Th. - jul97 TOC

Table of Contents: J. Approx. Theory, Volume 90, Number 1, July 1997

G. Mastroianni and J. Szabados
Bal\'azs-Shepard operators on infinite intervals, II
1--8

Yehoram Gordon, Shlomo Reisner, and Carsten Sch\"utt
Umbrellas and polytopal approximation of the Euclidean ball
9--22

Xiao Changbai and Li-He Zou
Reconstruction of band-limited functions from values on real
   sequences with an accumulation point
23--29

Ferenc Weisz
Ces\'aro summability of two-parameter trigonometric Fourier-series
30--45

Adelheid Fischer
On wavelets and prewavelets with vanishing moments in higher
   dimensions
46--74

Stefan Ericsson
Exact descriptions of some $K$ and $E$ functionals
75--87

G. G\'at
On the almost everywhere convergence of Fej\'er means of functions 
   on the group of $2$-adic integers
88--96

Vladimir N. Sorokin and Jeannette Van Iseghem
Algebraic aspects of matrix orthogonality for vector polynomials
97--116

A. I. Aptekarev, F. Marcell\'an, and I. A. Rocha
Semiclassical multiple orthogonal polynomials and the properties
   of Jacobi-Bessel polynomials            
117--146

Note

Paul Nevai and Ying Guang Shi
Notes on Steklov's conjecture in $L^p$ and on divergence of
   Lagrange interpolation in $L^p$
147--152



Date: Tue, 22 Jul 1997 12:24:01 -0500 (CDT)
From: "Prof. E.B. Saff" <esaff@tarski.math.usf.edu>
Subject: #13 Contents: Constructive Approximation, Volume 13, Number 4, 1997

Table of Contents:Constructive Approximation, Volume 13, Number 4, 1997

E. Kochneff
Expansions in Jacobi Polynomials of Negative Order
435--446

A. Kro\'{o}
On Approximation by Ridge Functions
447--460

S. Waldron
$L_p$-Error Bounds for Hermite Interpolation and the Associated Wirtinger
Inequalities
461--479

B. Beckermann and V. Kaliaguine
The Diagonal of the Pad\'{e} Table and the Approximation of the Weyl
Function of Second-Order Difference Operators
481--510

P.G.A. Floris and H.T. Koelink
A Commuting $q$-Analogue of the Addition Formula for Disk Polynomials
511--535

T.H. Koorwinder and A.L. Schwartz
Product Formulas and Associated Hypergroups for Orthogonal Polynomials on
the Simplex and on a Parabolic Biangle
537--567

T. Bloom and J.P. Calvi
Kergin Interpolants of Holomorphic Functions
569--583



Date: Thu, 7 Aug 1997 13:47:56 +0200
From: Baltzer Science <mailer@ns.baltzer.nl>
Subject: #14 Contents: Advances in Computational Mathematics 7-4

Advances in Computational Mathematics 7 (1997) 4

Tim N.T. Goodman, Charles A. Micchelli, Giuseppe Rodriguez and 
Sebastiano Seatzu
Spectral factorization of Laurent polynomials 429-454 

C.A. Micchelli and Thomas Sauer
Regularity of multiwavelets 455-545 

Y. Jeon, I.H. Sloan, E.P. Stephan and J. Elschner
Discrete qualocation methods for logarithmic-kernel integral equations on a
piecewise smooth boundary 547-571 

R. Hauber
Numerical treatment of retarded differential--algebraic 
equations by collocation methods 573-592 

Przemyslaw Bogacki, Stanley Weinstein and Yuesheng Xu
Distances between oriented curves in geometric modeling 593-621 



Date: Tue, 15 Jul 1997 11:41:44 -0400
From: Sergio Schuler <ss4@iprg.cise.ufl.edu>
Subject: #15 Answer: Construction almost tight frames (WD 6.7 #23)

Response:

>I want to construct a tight frame or a frame where the upper and lower
>bound are not far apart. I would prefer the frame to have an analytic
>expression in the time domain. 

The paper by Daubechies, "The wavelet transform, time-frequency
localization and signal analysis", in IEEE Transactions on Information
Theory, vol 36, no 5, Sept. 1990, has a good section on frame bounds
for arbitrary dilation and translation steps. She also shows the frame
bounds for several continuous functions (with analytic expressions in
the time domain). Her book also has details about the frame bounds and
more examples.

Question: What kind of wavelet do you need? Are there any requirements you 
would like to satisfy?

>I only want one scale however. Any help
>in doing this would be greatly appreciated. 

I guess that you don't want to have voices, is that correct?  Voices
usually make the frame tighter but add more redundancy. The tightness
of the frame basically depends on four things the dilation step ao,
the translation step bo, the number of voices and the function itself.

>Also information on
>computing frame bounds would be helpful. Specifically, computations
>such as the following: the Haar basis forms a tight frame when they
>are orthogonal and when they are used in a continuous wavelet
>transform, what happens for the case where the Haar functions are more
>redundant than the orthogonal case, how does one compute frame bounds
>for different redundancies?

The Haar basis are orthogonal when the dilation step ao=2 and the
translation step bo=1. If you change those values the Haar wavelet is
no longer orthogonal. However, the formulas by Daubechies will give
you the frame bounds.  A word of caution, if the frame bounds are not
tight and not equal to one that does not mean that the wavelet is not
orthogonal. Confused? The following is true: if the bounds are equal
and equal to one then the wavelet is orthogonal. The opposite however
is not true. The bounds are simply bounds.  They are not necessarily
the best bounds. So in some cases you might get bounds that do not
reflect how tight the frame really is. A very good example of this is
in the paper by Daubechies mentioned above (page 984, right column
after Theorem 2.9). The wavelet by Meyer mentioned there is
orthogonal, however, if one uses the wrong bounds it is impossible to
recognize that it is a frame. However, the modified bounds of Theorem
2.9 do show that the wavelet is indeed orthogonal.

With regards to computing the frame bounds, one needs to implement the
formulas given by Daubechies numerically. Although they look very
complicated the implementation is not too complicated. However, since
you are going to compute the bounds numerically, you must first prove
that the function you are using is indeed a frame. Again Daubechies
has a Theorem that shows what needs to be satisfied. Going back to the
numerical computation, since the sums are infinite one needs to
truncate them at some point. The approach that I use is to compute the
terms until they fall below the epsilon of the computer and then sum
them from smallest to biggest. This reduces the numerical error. Sup
and inf are of course replaced by max and min.  Also you only need to
compute the sums in the interval 1 to ao since the expressions are
"exponentially" periodic (i.e., the expressions for A and B as a
function of omega have the same behavior in 1 to ao, ao to ao^2, ao^2
to ao^3, ...). In the dyadic case 1 to 2, 2 to 4, 4 to 8, etc.

If you have any questions let me know.

Sergio



Date: Tue, 15 Jul 1997 09:55:01 -0400 (EDT)
From: unser@helix.nih.gov (Michael Unser)
Subject: #16 Answer: Question to the bio-medical world (WD 6.7 #19)

>I am beginning a PHD in the University of New South Wales Bio-medical
>department. I am interested in using wavelets for analysing ECG signals
>particularily for diagnostic information. Could anybody suggest papers and
>references that I can follow up?

Here are two general references on biomedical applications of wavelets
(including ECG) with many pointers to the literature.

[1]     A. Aldroubi and M. Unser, Wavelets in Medicine and Biology. Boca
Raton, FL: CRC, 1996.
[2]     M. Unser and A. Aldroubi, "A review of wavelets in biomedical
applications," Proceedings of the IEEE, vol. 84, no. 4, pp. 626-638, April
1996.



Date: Wed, 6 Aug 1997 11:46:14 -0600 (CST)
From: Jose de Jesus Perez Sevilla <josep@orion.gdl.cinvestav.mx>
Subject: #17 Answer: Wavelets in bio-world (WD 6.7 #19)

Dear Sir.

Here there are some articles in the subject of wavelts in ECG analisis.

-Li-Zheng-Tai,Detection of ECG Characteristic Points Using Wavelets 
Transforms, IEEE Transactions On Biomedical Engineering, vol 42, No 1 
January 1995. 

-Unser-Aldroubi, A Review of Wavelets in Biomedical 
Applications,Proceedings of the IEEE, Vol 84, No 4, April 1996.

-Sierra and others,Multirresolution Decomposition of the Signal Averaged 
ECG Using the Mallat Aproach for prediction of Arrhithmic events afer
Miocardial Infarction, Journal Of Electrocardiology, Vol 29 No 3, 1996

-Yang and Liao, Modeling and Decomposition of HRV Signals whit Wavelet 
Transforms, IEEE Engineering in Medicine and Biology MAGAZINE,Vol 16,
No 4, July/August, 1997.

-Metin Akay, Wavelet Applications in Medicine, IEEE Spectrum,Vol 34, No 5, 
May, 1997 

I hope this will be helpfull.

Regards,

Jose de Jesus Perez Sevilla
Centro de Investigacion y 
de Estudios Avanzados.(CINVESTAV)
Unidad Guadalajara. MEXICO.



Date: Thu, 31 Jul 1997 11:50:40 -0400
From: heller@aware.com (Peter Heller)
Subject: #18 Answer: Question on M-band wavelets (WD 6.7 #20)


There are a number of useful references and software packages for
M-band wavelets.  Our group at Aware has published a number of papers
in the EE and applied math literature; preprints can be found on our
ftp site ftp.aware.com via anonymous ftp, in the directory
/pub/papers/heller.  Closed-form expressions and algorithms for
constructing M-band wavelets can be found in the paper

P. N. Heller, 
{\em Rank $M$ Wavelet Matrices with $N$ Vanishing Moments,} 
SIAM J. Matrix Analysis, 16 (1995), pp. 502-518.

Useful earlier papers include:

P. Steffen, P. N. Heller, R. A. Gopinath, C. S. Burrus,
{\em Theory of regular $M$-band wavelets,} IEEE Trans. on Signal Processing,
41 (1993), pp. 3497-3511.

G. Welland and M. Lundberg, 
{\em Construction of compact $p$-wavelets}, 
Constructive Approximation, 9 (1993), pp. 347--370.

H. Zou and A. H. Tewfik, 
{\em Discrete orthogonal M-band wavelet decompositions}, in 
Proc. IEEE ICASSP-San  Francisco, March 1992.

J. Kautsky and R. Turkajova have also done some interesting work at Flinders
U. in Australia on the subject;  their work is available via ftp, as mentioned
in an old Wavelet Digest.

Aware, Inc. did at one time offer a software tool named WaveTool for
M-band wavelet design, but unfortunately that software is no longer
available.  However, you may find the Rice wavelet toolbox quite
useful in this regard.  It was developed by the Computational Math Lab
and DSP group at Rice, and can be found on the web at
http://www-dsp.rice.edu/software/RWT.  This is a set of Matlab and C
routines for designing M-band wavelets and performing the
corresponding wavelet transforms and generating sampled
representations of the continuous-time scaling and wavelet functions.
	As for applications to image compression, in addition to the Aware
papers available via ftp (e.g. Heller and Resnikoff, ICASSP '93), one
should consult H. Malvar's book {\em Signal Processing with Lapped
Transforms}, and the paper

R. L. de Queiroz, T. Q. Nguyen, and K. R. Rao,
{\em The GenLOT: generalized linear-phase lapped orthogonal transforms,}
IEEE Trans. on SP, 44 (1996), pp.497-507.

Good luck, I am always happy to answer further questions via email.

Peter Heller
Aware, Inc.
Bedford, MA 01730 USA
heller@aware.com



Date: Tue, 15 Jul 1997 12:40:13 +1000 (EST)
From: Richard Andrews <andrews@postoffice.utas.edu.au>
Subject: #19 Question: Non-separable wavelet reading 

Hi 

  I've recently moved into the field of Non-Separable Multidimensional
Wavelets and Filter Banks applied to image processing (particularly
quincunx and hexagonal sub-sampling lattices). I'm wonderring if
anyone out there is working in a similar field and could tell me what
they consider the best reading matter for someone new to the subject.

  All correspondence welcomed

      Thanks in advance

          Richard

Richard Andrews
University of Tasmania
Churchill Avenue, Sandy Bay
Hobart, Tasmania, AUSTRALIA
email     :  <andrews@postoffice.utas.edu.au>
phone (w) :  AUS +362 202 925



Date: Wed, 16 Jul 1997 11:14:29 +0100 (BST)
From: M J Darlington <ensmjd@bath.ac.uk>
Subject: #20 Question: Machine Duty Cycle Analysis using Wavelets

Hi, This tyro needs help!  As a small (but key) part of my research
into automating the design process I am trying to find a way of
extracting interesting features from the duty cycles of machines.

A duty cycle is a graphical representation of the force and velocity
requirement curves predicted for the machine (system) when performing
all its operations in a representative cycle of work.
        
I am motivated in this by the observation that human designers are
able to (and frequently do) use duty cycle curves to suggest the
initial solution to a design requirement.

The hypothesis is that, in some way, the information in the duty cycle
a) uniquely characterises the functional character of the system and
b) defining features in the duty cycles for a given system will be
recognisably similar to those of another system whose character (and
physical solution) are similar.  In essence I am trying to find a way
of automatically classifying as similar functionally similar systems,
based (in part) upon their duty cycle 'footprints'.

I have just stumbled upon wavelets (for the first time) as a possible
mechanism for extracting the necessary information. I have attempted
some of the introductory papers on the subject, but being a
non-mathematician have found them rather daunting. In short I am
unable to answer my question: "Are wavelets what I need?" They seem to
show promise (cope with aperiodicity, discontinous curves, and steep
gradients) but I am not sure whether they are in principal able to
cope with e.g. a) rather short data series (i.e. a short signal) and
b) the variation in ordering between the operations that would still
result in the same fundamental character in a system. Perhaps
answering my question will depend on a fuller description (I am very
happy to give this), but I would appreciate any response or
discussion.

Mansur Darlington, Research Officer
Engineering Design Centre in Fluid Power Systems
School of Mechanical Engineering
University of Bath
Claverton Down
Bath
BA2 7AY

e-mail M.J.Darlington@bath.ac.uk



Date: Wed, 16 Jul 1997 17:33:10 CDT
From: "MOHAN R. CHERUKURI" <MRC0291@CHENNOV2.TAMU.EDU>
Subject: #21 Question: On-line implementation of Wavelet Filtering.

Dear waveleteers,

I am working on using wavelets for filtering of noisy signals.  Most
of the literature gives methodology which works fine if I have a
signal of certain length.  My problem is, How can I use this filtering
technique on-line ? 'ie' as the signal keeps evolving how can I filter
it. Is there any way where in I can predict the next filtered output
based on the previous filtered outputs?

Any help on this matter including pointers to publications shall be
highly appreciated.

Thanks in advance...

Mohan.                 

MOHAN RAO CHERUKURI              Graduate Student,
4302, College Main #341,         Dept of Chemical Engg. ,
BRYAN, TX-77801                  Texas A&M University,
Ph # (409) 260-3846              College Station, Tx 77843.
                                 Ph # (409) 847-8713
               http://http.tamu.edu/~mrc0291



Date: Wed, 16 Jul 97 13:20:34 -0400
From: sghosh@rbubhop.mp.nic.in (Suddhasheel Ghosh)
Subject: #22 Question: Concerning lifting (WD 6.5 #21)

        This is in reference to WD 6.5 #21. You have mentioned a particular
technique called LIFTING. I am wondering whether this lifting technique comes
from Algebraic Topology where we can study about lifting of maps.

Best regards

Suddhasheel Ghosh, Consultant               e-mail :sghosh@rbubhop.mp.nic.in
M. P. Bhoj University,                      Ph:  91-0755-576555 (Off)
Red Cross Building, Shivaji Nagar
Bhopal. 462 011                             Fax: 91-0755-550606

---
Answer: No lifting does not have anything to do with lifting of maps
as known in Algebraic Topology. It is a general framework for
constructing wavelet transform, particularly on complicated
geometries. -- Wim Sweldens.



Date: Fri, 18 Jul 1997 10:02:07 +0200
From: Thomas <lich.imh@dbag.stg.daimlerbenz.com>
Subject: #23 Question: Which wavelet in which application ?

Hi there,
I'm a novice in using wavelets and so I have a general (maybe stupid
)question. Which wavelet could I use for which application. Maybe there
exists a rule for using wavelets in special applications, i.e.

        for De-Nosing a signal use sym4 or 
        for Identifying pure Frequencies use db3 
        ...

If there is no rule is it possible to get some information (Statistical
Information, i.e. mean-value, correlation ... ) about the signal which
says use these family of wavelets for this signal.

If anyone can help please contact me.

Thanks,

Thomas Lich
IMH

lich.imh@dbag.stg.daimlerbenz.com



Date: Sun, 20 Jul 1997 14:27:38 +0100
From: roger.fjortoft@cesbio.cnes.fr (Roger FJXRTOFT)
Subject: #24 Question: Edge detection in noisy images

Dear Wavelet Digest readers,

As a part of my thesis work, I try to evaluate the potential of
wavelet-based techniques for edge detection in noisy images, in
particular SAR images. Orthogonal wavelet transforms have been shown
to have excellent properties for image compression, but according to
several authors, shift-invariant (and thus redundant) transforms
should be used for image analysis problems such as edge detection. The
multiscale aspect is interesting in itself, but what exactly is the
contribution of wavelet theory in the case of edge detection based on
a shift-invariant transform?  Does the use of wavelets have any
advantage compared to other multiscale approaches, such as the Canny
edge detector or spatially adaptive operators?  Singularities are
characterized by their Lipschitz exponents, which can be estimated
from the evolution of the wavelet coefficients across the scales, as
described by Mallat et al. Has any robust, wavelet-based edge
detection algorithm been proposed, which explicitly states how the
results obtained on different resolution levels should be combined?
Thank you in advance!

Sincerely,

Roger Fjortoft
Centre d'Etudes Spatiales de la Biosphere (CESBIO) 
18 avenue Edouard Belin, bpi 2801, 31401 Toulouse cedex 4, France
Phone: (+33) 5 61 55 66 70
Fax: (+33) 5 61 55 85 00
E-mail: Roger.Fjortoft@cesbio.cnes.fr
Homepage: http://www.enseeiht.fr/Recherche/Info/Vision/Membres/RF/



Date: Fri, 25 Jul 1997 15:31:27 +0000 (GMT)
From: "S. Sriram" <ssram@protocol.ece.iisc.ernet.in>
Subject: #25 Question: Shiftability

 Hi Waveleters

 I'm a Masters student at Indian Institute of science.I work on
 Computer Vision.  In the July 97 issue of Pattern Recognition there
 was an article by KIM et al on "Stereo Matching Algo. Based on
 Modified Wavelet Decomposition Process"

 There they have over come an old problem of the shift invariance of
 the wavelet transform by not down sampling while taking the wavelet
 transform.They have showed with examples that it works.

 I would like to know why this works and will it always work?  If you
 have any suggestions or any reference materials,can you please
 forward them to my address.

 Thanking you ,
 Sriram Subramanian
 
 ssram@protocol.ece.iisc.ernet.in

[Answer]
[Updated answer]



Date: Sun, 27 Jul 1997 14:45:47 -0400
From: Ashok Ramasubramanian <ashok@dspse.com>
Subject: #26 Question: Wavelets in Speech Recognition

Hello All,
        I was wondering if there is any speech recognition
system/algorithm/method that uses wavelets. Any pointers to papers or
other technical material will be highly appreciated. 

Thank you. 

Ashok Ramasubramanian                Tel :(617)-275-3733 ext.230
Apt #302 .75 St.Aphonsus St,
Boston , MA-02120

e-mail:ashok@dspse.com
       aramasubrama@umassd.edu



Date: Fri, 1 Aug 1997 14:56:42 +0800
From: fabio@ned.dem.csiro.au (Fabio Boschetti)
Subject: #27 Question: Complex wavelets

Dear Readers,

I would be extremely grateful if you could send me basic references of
theory and applications of complex wavelets.

Thanks in advance

                                Fabio

Fabio Boschetti
email - fabio@ned.dem.csiro.au

[Answer]



Date: Fri, 08 Aug 1997 08:26:21 +0700
From: Danny Salim <dannysalim@usa.net>
Subject: #28 Question: Pitch detection in music signal

I am an undergraduate who has been working on my thesis about detecting
pitch in music signal using wavelet.
I'd like to know what is the best method to do it.



Date: Sun, 10 Aug 1997 10:16:51 -0400 (EDT)
From: Hamid Soltanian-Zadeh's Graduate Students <zadeh@rad.hfh.edu>
Subject: #29 Question: Looking for a public domain software  

Dear Readers

 I am looking for a 2d wavelet software for Sun solaries
for image processing. I will be happy if anyone help me.

Reza Nezafat
zadeh@nutbush.rad.hfh.edu



Date: Mon, 11 Aug 1997 19:00:44 +0800
From: Ma Feng <fma@cs.hku.hk>
Subject: #30 Question: 3D wavelet software.

Hi, 

  Now, there are many wavelet softwares dealing with 1D and 2D wavelet
transform including discrete and continuous versions.  

  Is there any software dealing with 3D wavelet transform?

  Thanks,
-- 
Ma Feng 
Dept. of Computer Science       Graphics Lab: CYC415 (852)2857 8454
The University of Hong Kong     Office:       HW520  (852)2857 8273
Pokfulam Road, Hong Kong        Email To:     fma@cs.hku.hk


<webmaster@wavelet.org>