[Home]
[Back Issues]
[Search]
[Calendar]
[Vol 1]
[Vol 2]
[Vol 3]
[Vol 4]
[Vol 5]
[Vol 6]
[Vol 7]
The Wavelet Digest, Volume 7, Issue 3 (March 17, 1998)
Today's Editor: Wim Sweldens
Lucent Technologies, Bell Laboratories
wim@bell-labs.com
Today's Topics:
1: News: Journal of Fourier Analysis changes publisher
2: Preprint: Unsubsampled Wavelet Transform and the Lifting Scheme
3: Preprint: Preprint from Tim Downie
4: Preprint: Signal and Image Denoising
5: Preprint: Biorthogonal Multiwavelets Using the Lifting Scheme
6: Preprint: Optimal ON filter banks for multiple description coding
7: Preprint: Wavelet analysis of the solar neutrino capture data
8: Web: IMI Reports from Univ. of South Carolina
9: Course: Multiscale approaches for partial differential equations
10: Course: Compression, Wavelets, Neural Networks, Fuzzy Logic, Chaos
11: Job: Research Associate position in Adelaide
12: Contents: Numerical Algorithms 16 1-2
13: Contents: J. Approx. Th. - feb98 TOC
14: Answer: Image Compression programs using Wavelet (WD 7.2 #20)
15: Answer: Wavelets and integral equations in 2 and 3D (WD 5.2 #28)
16: Answer: Wavelets and integral equations in 2 and 3D (WD 5.2 #28)
17: Answer: Modulus and phase of Morlet wavelet using FFT (WD 7.2 #19)
18: Answer: Realtime fast wavelet transform (WD 7.2 #14)
19: Question: parametrization of wavelets
20: Question: Wavelet analysis of medical audio signals
21: Question: IIT looking for wavelet material
22: Question: SAS wavelet technique
23: Question: Looking for papesr on Discrete Wavelet Multi-Tone (DWMT)
24: Question: Tracking phase information
25: Questions: Wavelets and analysing EMG signals
26: Question: Reinvented wavelets?
Current number of members: 11246
E-mail:
General 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
Receive publishing policy: policy@wavelet.org
Visit us on the Web at http://www.wavelet.org/
Calendar of events:
Mar 18-20: CISS '98 Call For Papers, Princeton, NJ
Apr 13-17: International Wavelet Conference, Tanger, Morocco
*Jun 1-12: Multiscale approaches for PDE, Marseilles, France
*Jun 8-12: Wavelets, Neural Networks, Fuzzy Logic, Chaos, UCLA
Jun 22: IEEE Multimedia compression, Santa Barbara, CA
Jun 29-Jul 10: EMS Summer School on Wavelets, Paris
Jul 8-10: Nonlinear Modeling, Leuven, Belgium
Jul 30-Aug 7: Workshop on Self-Similar Systems, Dubna, Russia
Oct 4-9: School on Wavelets in the geosciences, Delft, Netherl.
Oct 5-8: 10th GAMM-Workshop on Multigrid Methods, Bonn, Germany
Date: Fri, 6 Mar 1998 17:36:35 -0500
From: "Wayne Yuhasz" <yuhasz@birkhauser.com>
Subject: #1 News: Journal of Fourier Analysis changes publisher
Announcement
Birkhauser Boston Acquires "Journal of Fourier Analysis and
Applications" from CRC Press
Recently Birkhauser Boston completed arrangements for assuming all
responsibilites for the Journal of Fourier Analysis and Applications,
effective January 31, 1998. John Benedetto remains the
editor-in-chief and the editorial office is still at the Univerity of
Maryland, under John's supervision. All current editorial procedures
and committments will remain in effect and continue
uninterrupted. The journal will retain the same physical appearance,
frequency of publication, and subscription price.
The Journal of Fourier Analysis and Applications joins a very select
and distinguished group of mathematics journals from Birkhauser
world-wide. We are very pleased and excited to have the journal and
its distinguished editorial board as part of Birkhauser publications.
During the transition period , the points of contact at Birkhauser
Boston are Wayne Yuhasz <yuhasz@birkhauser.com> for editorial matters
or Karen Gray for marketing and distribution matters
<gray@birkhauser.com >. North American subscription fulfillment
<matters are handled by Springer-Verlag New York
<<custserv@springer-ny.com> or 1-800-Springer. For all queries
<concerning distribution and subscriptions outside
North America contact Birkhauser Basel , Sabine Voss <voss@birkhauser.ch>.
Wayne Yuhasz
Executive Editor
Computational Sciences & Engineering
yuhasz@birkhauser.com
Birkhauser Boston
fax 617-876-1272
www.birkhauser.com
Date: Fri, 20 Feb 98 11:12:26 +0100
From: "Alexander Stoffel" <stoffel@nebsy.nt.fh-koeln.de>
Subject: #2 Preprint: Unsubsampled Wavelet Transform and the Lifting Scheme
Title: Remarks on the Unsubsampled Wavelet Transform and the Lifting Scheme
Author: Alexander Stoffel, Fachhochschule Koeln,
Fachbereich Nachrichtentechnik, Betzdorfer Strasse 2,
D-50679 Koeln, Germany,
e-mail:
stoffel@nebsy.nt.fh-koeln.de
Abstract:
The lifting scheme, which is known as a very useful tool for the
wavelet transform, is adapted to the calculation of the unsubsampled
wavelet coefficients. It is shown that - as in the subsampled case -
the two-band transform for each FIR filter pair without common zeros
can be performed by a finite number of lifting steps. Inverting each
step, one gets a perfect reconstruction equivalent to the arithmetic
mean of the result of two different reconstruction filters. Because of
this easy inversion, the lifting scheme offers a great flexibility
treating boundary conditions in the case of finitely many data.
Keywords: Wavelet, filter bank, lifting scheme, finite length signals
Status: Submitted to Signal Processing
The preprint is available as compressed PostScript or PDF file at
http://www.fh-koeln.de/fb/fb-nt-inf/stoffel/lifting.html
A previous (longer) version is also available from this web site.
Alexander Stoffel
Fachhochschule Koeln
Fachbereich Nachrichtentechnik
e-Mail: stoffel@nebsy.nt.fh-koeln.de
http://www.fh-koeln.de/fb/fb-nt-inf/stoffel/stoffel.html
Tel.:(49) 221 830 15 83
Fax :(49) 221 82 75 24 45
Date: Thu, 05 Mar 1998 14:37:09 +1100
From: Tim Downie <tim.downie@cmis.csiro.au>
Subject: #3 Preprint: Preprint from Tim Downie
The following are available in postscript format from my
web page http://www.dms.csiro.au/~timd/research.html
Thanks for your interest
Tim
1) "Wavelets in statistics" PhD Thesis Downie T.R.
University of Bristol, 1997.
Abstract
Two different aspects of wavelet methods in statistics are studied,
preceded by a review of existing methods. The first aspect is
thresholding using multiple wavelets, and the second is fitting a
wavelet basis to the deformable template model.
A multiple wavelet basis is a wavelet basis generated by $L \geq 2$
mother wavelets. The discrete multiple wavelet transform DMWT
transforms a set of scalar data into a set of $L-$vector
coefficients. In order to obtain an efficient multiwavelet
decomposition the data must be preprocessed first. Four criteria
are developed to assess different preprocessing methods. A
multivariate thresholding method is proposed which adapts each
coefficient vector as one entity. This accommodates the
time-frequency properties of the discrete multiple wavelet
transform, and the correlation within vector coefficients. Hard
and Soft thresholding rules, and a universal threshold are
developed. Results from a simulation study demonstrate that
multivariate thresholding gives a lower mean square error than
using univariate thresholding of multiple wavelets and thresholding
using Daubechies wavelets with the universal threshold. Using
different preprocessing methods gave markedly different results.
The method of deformable templates is used to deform one
image(template) into the shape of another similar image. The
template is deformed by a deformation function. This deformation
function is often modelled using a Fourier basis so that a
deformation can be obtained numerically. A wavelet basis should be
better at representing deformations that have localised features.
A data set where the deformation is likely to be localised is a
collection of femoral condyle images. A wavelet model using a
mixture distribution is proposed and a penalised least squares
method is developed in order to obtain a deformation between a
template and image. Three algorithms are developed and compared by
fitting a deformation between two femoral condyle images.
2) "Signal Preprocessing for Multiwavelets" Downie T.R.
Abstract:
Orthogonal wavelet bases have been developed using multiple mother
wavelet functions. Applying the discrete multiple wavelet
transform requires the input data to be preprocessed to obtain a
more economical decomposition. Four general properties of
preprocessing methods are reviewed: length, degree, orthogonality
and frequency response. A minimal prefilter is defined to have
length one and highest attainable degree. For the GHM Multiwavelet
specific prefilters are discussed, including a minimal matrix
prefilter and a repeated signal prefilter. The criteria for each
prefilter are obtained and how this compares to the performance of
prefilters.
3) "The Discrete Multiple Wavelet Transform and Thresholding
Methods." Downie T.R. & Silverman B.W. (revised version)
abstract:
We propose thresholding for multiwavelets considering the
coefficient vectors as a whole rather than thresholding individual
elements. A multivariate universal threshold is obtained using the
$\chi ^2$ distribution. Simulations indicate that, using the GHM
multiwavelet with appropriate preprocessing, our method outperforms
univariate thresholding of both GHM and Daubechies wavelet
decompositions.
4) "Economical Representation of Image Deformation Functions
Using a Wavelet Mixture Model"
Downie T.R. Shepstone L.& Silverman B.W.
Abstract: One method of analysing two similar images is to obtain a
deformation function that maps one image onto the other. The
modelling of this deformation function in terms of a functional
basis leads to numerical methods of obtaining a good deformation.
It is advantageous to be able to represent a wide class commonly
observed deformations economically, ie where most coefficients are
zero. We proposed a wavelet model for the deformation, where each
wavelet coefficient has the mixture distribution $\pi \delta _0 +
(1-\pi) N(0,\sigma ^2)$. This distribution reflects our prior
belief in the wavelet coefficients and results in an economical
representation for the deformation. To implement this method, a
penalised least squares methodology is adopted and three algorithms
are devised. A numerical assessment of the method is made by
applying the algorithms to some femoral condyle images, which
require the deformations to have localised features.
--
Tim Downie
Analysis of Large and Complex Datasets Project
CSIRO Mathematical and Information Sciences
Locked Bag 17, North Ryde, NSW 2113, Australia
Phone: +61 2 9325 3256 Fax: +61 2 9325 3200
email:tim.downie@cmis.csiro.au
url:http://www.dms.csiro.au/~timd/research.html
Date: Tue, 10 Mar 1998 19:26:34 -0500
From: Vasily Strela <strela@pascal.Dartmouth.EDU>
Subject: #4 Preprint: Signal and Image Denoising
Signal and Image Denoising via Wavelet Thresholding:
Orthogonal and Biorthogonal, Scalar and Multiple Wavelet Transforms
by V. Strela (Dartmouth College, NH, USA, strela@cs.dartmouth.edu)
and A. T. Walden (Imperial College, London, UK, a.walden@ic.ac.uk)
Imperial College, Statistics Section, Technical Report TR-98-01
Abstract
In this paper we discuss wavelet thresholding in the context of scalar
orthogonal, scalar biorthogonal, multiple orthogonal and multiple
biorthogonal wavelet transforms. Two types of multiwavelet
thresholding are considered: scalar and vector. Both of them take into
account the covariance structure of the transform. The form of the
universal threshold is carefully formulated. The results of numerical
simulations in signal and image denoising are presented.
Multiwavelets outperform scalar wavelets for three out of four noisy
1D test signals, and the Chui-Lian scaling functions and wavelets
combined with repeated row preprocessing appears to be a good general
method. Vector thresholding does not always outperform scalar
thresholding.
Multiwavelets generally outperform scalar wavelets for image denoising
for all four noisy 2D test images, and the results are visually very
impressive. Only for `Lenna' and `fingerprints' with signal to noise
ratios of 2 do scalar wavelets perform best. As for 1D signal
processing, Chui-Lian scaling functions and wavelets combined with
repeated row preprocessing appears to be a good general method.
For both 1D and 2D cases, the reconstructed signals derived from such
a good general method demonstrate much reduced noise levels ---
typically 50\% of the standard deviation of the original noise.
Tha paper can be downloaded from
http://math.dartmouth.edu/~strela/
or
http://www.ma.ic.ac.uk/~atw/
Date: Wed, 11 Mar 1998 12:32:53 +0800 (SST)
From: Goh Say Song <matgohss@leonis.nus.edu.sg>
Subject: #5 Preprint: Biorthogonal Multiwavelets Using the Lifting Scheme
Preprint: Construction of Biorthogonal Multiwavelets Using the Lifting
Scheme
Authors: Say Song Goh, Qingtang Jiang and Tao Xia
Department of Mathematics
National University of Singapore
10 Kent Ridge Crescent
Singapore 119260
E-mail: Say Song Goh (matgohss@leonis.nus.edu.sg)
Qingtang Jiang (qjiang@haar.math.nus.edu.sg)
Tao Xia (xiatao@haar.math.nus.edu.sg)
Abstract:
The lifting scheme has been found to be a flexible method for
constructing scalar wavelets with desirable properties. Here it is
extended to the construction of multiwavelets. It is shown that any
set of compactly supported biorthogonal multiwavelets can be obtained
from the Lazy matrix filters with a finite number of lifting steps.
Based on the proposed lifting scheme, multiwavelet transforms that map
integers to integers are obtained. As an illustration of the general
theory, compactly supported biorthogonal multiwavelets with optimum
time-frequency resolution are constructed. In addition, experimental
results of applying these multiwavelets to image compression are
presented.
Date: Wed, 11 Mar 1998 20:13:17 -0600 (CST)
From: Xuguang Yang <yangx@ifp.uiuc.edu>
Subject: #6 Preprint: Optimal ON filter banks for multiple description coding
Dear Waveleters,
We would like to inform you the availability of a preprint of our
recent work on multiple description coding using orthonormal filter
banks. It is going to be presented at CISS-98 (princeton), and a
submission to IEEE Trans. on Signal Processing is in preparation.
Title: Redundancy rate-distortion optimal orthonormal filter banks for
multiple description coding
Authors: Xuguang Yang and Kannan Ramchandran
Abstract: Multiple description coding refers to encoding an
information source into multiple bitstreams, in a way that various
levels of tradeoff between decoding quality and channel protection can
be supported. This paper addresses the design of two-channel
orthonormal filter banks to optimize the redundancy rate-distortion
performance of multiple description subband coding. The problem is
formulated in a Lagrangian optimization framework, and is solved
directly in the frequency domain to produce the optimal filter
responses. By varying the Lagrangian parameter, our design obtains all
points on the information redundancy rate-distortion curve for
Gaussian wide sense stationary input processes. Application of the
proposed algorithm to the popular AR(1) model yields very interesting
and insightful results. Comparisons with popularly deployed filter
banks reveal large gaps from the theoretical performance bound.
http: //www.ifp.uiuc.edu/~yangx/research
Date: Mon, 16 Mar 1998 14:45:50 EST
From: Hans J. Haubold <haubold@ekpvs2.dnet.tuwien.ac.at>
Subject: #7 Preprint: Wavelet analysis of the solar neutrino capture data
Wavelet analysis of the new solar neutrino capture rate data for the
Homestake experiment
Author: H.J. Haubold
Abstract: Wavelet analysis offers an alternative to Fourier based
time-series analysis and is particularly useful when spectral features
are time dependent. We analyze the solar neutrino capture rate of the
radiochemical Homestake chlorine experiment with abbreviated Morlet
wavelets, using Foster's (Astronomical Journal, 111, 1709(1996))
rescaled wavelet technique. We emphasize the complementarity of
wavelet analysis to Fourier analysis. Wavelet analysis confirms the
results of previously undertaken Fourier and Lomb-Scargle
analysis. The Homestake data seem to contain a harmonic content with
periodicities of 4.76 yr, 1.89 yr, 0.85 yr, and 0.51 yr. Wavelet
analysis reveals that the 4.76 yr and 1.89 yr periods show an almost
constant behavior over the 25 yr Homestake data record, while the 0.85
yr and 0.51 yr periods exhibit a transient phenomenon. The analysis
does not show strong evidence for a period of the solar 11 yr cycle.
Status: To be published in the proceedings of the Workshop on Data
Analysis Techniques, Brazil, November 1997.
LaTeX and .ps files, including 6 gif figure files, are available at
http://xxx.lanl.gov/abs/astro-ph/9803136
Date: Thu, 5 Mar 1998 14:40:36 -0500 (EST)
From: Susanne Brenner <brenner@math.sc.edu>
Subject: #8 Web: IMI Reports from Univ. of South Carolina
This is to announce the on-line availability of the
Industrial Mathematics Initiative (IMI) Research Reports:
http://www.math.sc.edu/~imip/
This is a preprint series which contains reports by the faculty and
visitors of the Department of Mathematics at the University of South
Carolina.
Date: Tue, 10 Mar 98 12:14:17 GMT
From: liandrat@marius.univ-mrs.fr (Jacques Liandrat)
Subject: #9 Course: Multiscale approaches for partial differential equations
EUROPEAN SUMMER SCHOOL
Multiscale approaches for partial differential
equations
Marseilles, France, June 01 - 12 1998
Presentation:
Multiscale approaches are widely developped in various domains related
to partial differential equations such as functional analysis,
numerical analysis or numerical algorithms.
The School should provide a global presentation of the different
approaches, allowing interactions and comparisons between them. One
of the main goal of the School is to emphasize the connections between
theory and application in the approximation of partial differential
equations.
Preliminary Program:
The school will be organized in the form of lectures, computer sessions
and seminars.
The main topics of the school will be:
1) Functional spaces and non-linear approximation
2) Wavelets, approximation and numerical algorithms
3) Multigrid methods and domain decomposition technics
4) Spectral methods and non-linear Galerkin methods
5) Multilevel methods for conservation laws
6) Implementation on computers
The lecturers will be:
Silvia BERTOLUZZA (CNR, University of Pavia, Pavia (Italy))
Claudio CANUTO (Politech di Torino, Torino (Italy))
Albert COHEN (Universite Pierre et Marie Curie, Paris (France))
Wolfgang DAHMEN (IGPM, Aachen (Germany))
Rosa DONAT (University of Valencia, Valencia (Spain))
Francois JAUBERTEAU (Universite Blaise Pascal, Clermont Ferrand (France))
Jacques LIANDRAT (ESM2/IRPHE, Marseille (France))
Philippe TCHAMITCHIAN (LATP, Universite Aix Marseille III, Marseille (France))
Each participant is expected to present a short seminar or a poster.
Registration and address:
School Secretary:
Francoise CESSOT-TAVERA
ESM2, IMT, Technopole de Chateau-Gombert,
13451 Marseille Cedex 20
Fax: 04-91-05-45-98
Phone: 04-91-05-44-11
E-mail: ess@esm2.imt-mrs.fr
Registration:
The number of participant is limited to 30 and a selection will be
made, if necessary. The registration fees are fixed to 1500FF. The
price of accomodation at the Centre de Formation de Luminy is fixed to
354FF per day (including room, breakfast, lunch and dinner).
Funding is available for students of the European Community whose
University is engaged in the ERASMUS Intensive Program ``Multiscale
approaches for partial differential equations''. Persons who need
fundings to attend the School should contact the School secretary.
Registration forms are available from:
http://esm2.imt-mrs.fr/ess/
Organizing Committee:
J. Liandrat
ESM2/IRPHE, Universite d'Aix-Marseille II, IMT, Technopole de
Chateau-Gombert, 13451 Marseille Cedex 20, France e-mail:
liandrat@marius.univ-mrs.fr
Ph. Tchamitchian,
Laboratoire de Mathematiques Fondamentales et Appliquees, Universite
d'Aix-Marseille III, Faculte de St Jerome, Avenue Escadrille
Normandie-Niemen, case cour A, 13397 Marseille cedex 20, France.
e-mail: tchamphi@math.u-3mrs.fr
Date: Tue, 10 Mar 1998 09:44:09 -0800
From: "Goodin, Bill" <bgoodin@unex.ucla.edu>
Subject: #10 Course: Compression, Wavelets, Neural Networks, Fuzzy Logic, Chaos
On June 8-12, 1998, UCLA Extension will present the short course,
"Video/Image Compression, Wavelets, Neural Networks, Fuzzy Logic, and
Chaos", on the UCLA campus in Los Angeles.
The instructors are Harold Szu, PhD, Lamson Professor of Computer
Science, University of Southwestern Louisiana, Lafayette, and Past
President of the International Neural Network Society (INNS); and
Charles Hsu, PhD, Principal Engineer, Trident Systems.
While the disciplines of wavelets, neural networks, fuzzy logic, and
chaos seem divergent, their underlying principles converge for the
purpose of solving invariant pattern recognition problems. For
example, the complex neurodynamic patterns in human visual and hearing
systems can be related to video redundancy reduction which generates
invariant features for detection, classification, and identification
purposes.
The course describes various applications in chaos, fuzzy logic, ANN,
WT, and associated learning algorithms, illustrated in terms of
spatio-temporal information processing such as signal/image de-noise;
signal classification; control device/machine chaos; video-phone
communication codec; chaotic biomedical and heart applications;
intelligent biosensors such as pacemakers and hearing aids; bomb
sniffers, and discrete wavelet transform subband coding for image and
motion estimation for video compression. Participants see how the
human visual system is organized similar to the WT of the
Gabor/Morlet-like edge-map pyramid in terms of the Multiple Resolution
Analyses (MRA) mathematics which reduce the spatial and temporal
redundancies for invariant pattern recognition.
UCLA Extension has presented this highly successful short course
since 1993.
The course fee is $1495, 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: Fri, 27 Feb 98 14:47:30 +1030
From: "J. Kautsky" <jarka@ist.flinders.edu.au>
Subject: #11 Job: Research Associate position in Adelaide
FLINDERS UNIVERSITY OF SOUTH AUSTRALIA
FACULTY OF SCIENCE AND ENGINEERING
DEPARTMENT OF MATHEMATICS AND STATISTICS
RESEARCH ASSOCIATE
(Fixed-term)
$31,722 - $42,621 pa
(Minimum starting salary for appointee with PhD will be $39,787 p.a.)
Ref 98045 Applications are invited for a 1 year appointment for an ARC
funded position concerned with research on "Design of Task-Adapted
Generalised Wavelets". The emphasis of the research is both on the
theory, particularly properties and representation of both orthogonal
and biorthogonal multi-band wavelets and multi wavelets, as well as
development of directly applicable algorithms.
The appointee will work with a research group including Professor
W. Moran, Dr. J. Kautsky, research students and visitors. Excellent
computing and library facilities are available.
Essential criteria includes a PhD or significant progress towards
completion of a PhD in mathematics or in an area with strong
mathematical content; demonstrated ability to conduct research;
interpersonal skills; and computing skills (preferably UNIX, MATLAB
and C).
Appointment would be for 1 year in the first instance, with possible
re-appointment for up to nine months dependent on availability of
funds. Some assistance towards relocation expenses (if applicable)
may be negotiated.
Commencement date preferably before mid-1998. For further information
and copies of the selection criteria please contact Prof. W. Moran on
telephone (+61)8 8201 2712 or email bill@ist.flinders.edu.au or Dr. J.
Kautsky on telephone (+61)8 8201 2360 or email
jarka@ist.flinders.edu.au.
Conditions of appointment will be specified in any offer of
appointment which may be made as a result of this advertisement.
Applicants must address the selection criteria in the documentation.
Applications, quoting the reference number, and giving qualifications,
experience and the names, addresses (including Email), and facsimile
numbers of three referees of whom confidential enquiries may be made,
should be lodged in duplicate with the Manager, Human Resources,
Flinders University, GPO Box 2100, Adelaide SA 5001, or fax (08) 201
3131 by 27 March, 1998.
The University reserves the right not to make an appointment or to invite
applications. Applications from suitably qualified women are welcome.
The University is an equal opportunity employer and promotes a smoke-free
work environment.
Date: Tue, 3 Mar 1998 16:27:23 +0100
From: Baltzer Science <mailer@ns.baltzer.nl>
Subject: #12 Contents: Numerical Algorithms 16 1-2
B. Dumitrescu, M. Doreille, J.-L. Roch and D. Trystram
Two-dimensional block partitionings for the parallel sparse Cholesky
factorization 17-38
J.P. Gregoire, C. Rose and B. Thomas
Direct and iterative solvers for finite-element problems 39-53
Vincent Heuveline, Bernard Philippe and Miloud Sadkane
Parallel computation of spectral portrait of large matrices by Davidson type
methods 55-75
Gerard Meurant
The computation of bounds for the norm of the error in the conjugate
gradient algorithm 77-87
M. Mitrouli and C. Koukouvinos
On the computation of the Smith normal form of compound matrices 95-105
W.H. Enright and Min Hu
Improving performance when solving high-order and mixed-order boundary value
problems in ODEs 107-116
Relja Vulanovic
Fourth order algorithms for a semilinear singular perturbation problem
117-128
Qiming He, Lishan Kang and D.J. Evans
Convergence and stability of the finite difference scheme for nonlinear
parabolic systems with time delay 129-153
Tim Gutzmer and Armin Iske
Detection of discontinuities in scattered data approximation 155-170
Yves Lucet
Faster than the Fast Legendre Transform, the Linear-time Legendre Transform
171-185
Luc Rebillard
Rational approximation in the complex plane using a \tau-method and computer
algebra 187-208
J.C. Butcher, P. Chartier and Z. Jackiewicz
Nordsieck representation of DIMSIMs 209-230
Date: Tue, 10 Mar 1998 10:28:06 -0500 (EST)
From: Paul Nevai <nevai@math.ohio-state.edu>
Subject: #13 Contents: J. Approx. Th. - feb98 TOC
Table of Contents: J. Approx. Theory, Volume 92, Number 2, February 1998
George Kvernadze
Determination of the jumps of a bounded function by its Fourier series
167--190
G. G\'at
On the two-dimensional pointwise dyadic calculus
191--215
M. Bello Hern\'andez and G. L\'opez Lagomasino
Ratio and relative asymptotics of polynomials orthogonal on an arc of
the unit circle
216--244
Will Light and Henry Wayne
On power functions and error estimates for radial basis function
interpolation
245--267
Riadh Ben Ghanem and Cl\'ement Frappier
Explicit quadrature formulae for entire functions of exponential type
267--279
Andrei Mart\'{\i}nez-Finkelshtein, Juan J. Moreno-Balc\'azar,
Teresa E. P\'erez, and Miguel A. Pi\~nar
Asymptotics of Sobolev orthogonal polynomials for coherent pairs of
measures
280--293
Jos\'e J. Guadalupe and Mario P\'erez
Perturbation of orthogonal Fourier expansions
294--307
A. L. Levin, V. V. Maimeskul, and E. B. Saff
Rational approximation with locally geometric rates
308--330
E. A. Rakhmanov, E. B. Saff, and P. C. Simeonov
Rational approximation with varying weights, II
331--338
Table of Contents: J. Approx. Theory, Volume 92, Number 1, January 1998
Thomas Bloom
Some applications of the Robin function to multivariable approximation
theory
1--21
M. Kazim Khan, B. Della Vecchia, and A. Fassih
On the monotonicity of positive linear operators
22--37
Mihai Putinar
Extremal solutions of the two-dimensional $L$-Problem of moments, II
38--58
Ryszard Szwarc
Chain sequences, orthogonal polynomials, and Jacobi matrices
59--73
Ulrike Maier
On best approximation of the monomials on the unit ball of $\BR^r$
74--81
Bernd Mulansky and Marian Neamtu
Interpolation and approximation from convex sets
82--100
M. Bartelt and W. Li
Characterization of generalized Haar spaces
101--115
R. Sakai
Lagrange interpolation based at the zeros of orthonormal polynomials
with Freud weights
116--127
J. A. Oram and V. Davydov
Best approximation by periodic smooth functions
128--166
Date: Tue, 17 Feb 1998 16:37:45 +1100
From: Andrew Dorrell <andrewd@ee.uts.edu.au>
Subject: #14 Answer: Image Compression programs using Wavelet (WD 7.2 #20)
Roberto writes:
> I am intereset in wavelet based image compression programs. After I
> read some articles, I would like to know where I can find complete
> algorithms and source programs written in "C" for this applications.
There is a Wavelet Image Compression Construction Kit which should fit
much of your needs at
http://www.cs.dartmouth.edu/~gdavis/wavelet/wavelet.html
Date: Tue, 17 Feb 1998 16:50:32 +1100
From: Andrew Dorrell <andrewd@ee.uts.edu.au>
Subject: #15 Answer: Wavelets and integral equations in 2 and 3D (WD 5.2 #28)
Ingve Simonsen wrote:
>
> Dear Andrew
>
> In the Wavelet Digest 1996 you had the following question:
>
> >Dear waveleters,
> >
> >In G. Beylkin, R. Coifman and V. Rokhlin, "Fast Wavelet Transforms and
> >Numerical Algorithms I.", the authors describe techniques for applying
> >operators to one dimensional data in the wavelet domain. They also
> >note that extension to higher dimensions is relatively
> >straightforward.
> >
> >I am interested in knowing if anyone has implemented these techniques >in
> >2D or 3D. Alternatively, what literature might be available describing
> >how this could be done?
> >
> >Thanks in advance,
> >
> >Mr Andrew Dorrell
> >email: andrewd@ee.uts.edu.au
>
> I an curious what kind of answers (if any) you got. Any references codes
> etc.
The only answers I got were ones like your own :-( So I had to work
out the details myself and they form a component of my thesis which is
now being vetted prior to final submission.
In short the computational characteristics are pretty nasty when you
use the regular ND transform structure which is separable only over
one level. Of course they are similar to the BCR case when the N
level transform is applied to rows then columns of the data (ie. in
completely separable fashion)
If you are interested in more detail I can find out what the
publication policy of UTS is for theses --- ie how soon I will be able
to publish my thesis globally. I can't really say whether I will be
publishing a paper covering the work at this stage.
Date: Mon, 16 Mar 1998 10:34:09 -0800
From: Peter Schroeder <ps@cs.caltech.edu>
Subject: #16 Answer: Wavelets and integral equations in 2 and 3D (WD 5.2 #28)
Andrew,
Looking over the old WDs I found your question. This may be a bit
late, but you can find examples of fast wavelet techniques for the
integral equations which govern radiative heat transfer and
radiosity. Basically a Fredholm Integral equation of the second kind
with a kernel satisfying the needed Calderon-Zygmund bounds. The
domain is 2 dimensional since the relevant quantities are defined as
functions of surfaces. There are also 4D versions when scattering has
a directional dependence. Wavelet algorithms for radiosity and
radiance have been quite an active research area in computer
graphics. The techniques are very fast indeed and researchers have
reported running multi-million element simulations on workstation
class machines. If you want to learn more about this, search for
``Hierarchical Radiosity'' as a key. The book by Cohen and Wallace
(``Radiosity and Realistic Image Synthesis'') has a chapter on this
and an extensive bibliography. Other good sources of research on this
are the proceedings from the Eurographics Workshop on Rendering which
are published by Springer (most recent: ``Rendering Techniques 97'').
Peter
Date: Thu, 19 Feb 1998 11:01:37 +0300
From: tap@meteo.ru
Subject: #17 Answer: Modulus and phase of Morlet wavelet using FFT (WD 7.2 #19)
>Please, can anybody explain: Is it possible to divide modulus and
>phaze while calculating Morlet wavelet with use available FFT and
>IFFT procedures (for instance in MATHCAD)? I can calculate modulus
>and phaze throuh direct integrating but when I use these procedures I
>get a single 1-dimension set and cannot get modulus and phaze
>differently. May be it is impossible with completed FFT-IFFT
>procedures and it is necessary to use only algoritms?
Hello Anna, I have seen your question on the wavelet digest. I am
presently using the Morlet wavelet to analyse turbulence measurements
coming from a hot plasma, so I can perhaps try to help you. First of
all, I don't know how the FFT is implemented in Mathcad. If well
implemented, it should give a complex result. In general, the
procedure for calculating the wavelet transform of a time series x(t)
through FFT is to compute the FFT of x(t), to compute the FFT of the
wavelet w(t), to multiply them and to compute the anti-FFT. Now, the
FFT of x(t), which I'll call X(f), is complex, with the property that
the negative frequency compnents are complex conjugates of the
positive ones, since x(t) is real, i.e. X*(-f)=X(f) [I use the * to
indicate the complex conjugation]. The FFT of the Morlet wavelet
w(t), which I'll call W(f), is real, and does not have simmetry
between negative and positive frequencies, i.e. W(-f) is not equal to
W(f). The product of the two, P(f)=X(f)W(f), is therefore complex,
and does not possess the same property as the FFT of x, i.e. IT IS NOT
TRUE that P*(-f)=P(f). As a consequence, the anti-FFT of P(f) is
complex, and you can therefore compute a modulus and a phase. A
warning: are you sure that the phase is so meaningful? I have been
thinking about this. The wavelet transform is localised in time, and
the phase simply gives a fine adjustment of this localisation, by an
amount less than the distance between 2 subsequent points of the
transform. In my applications, this is seldom important, but may be in
your case it is different. I hope to have been helpful. Contact me if
you need further clarifications. Yours.
Emilio
Emilio Martines E-mail: martines@igi.pd.cnr.it
Consorzio RFX phone: +39 (49) 8295065
corso Stati Uniti, 4 fax: +39 (49) 8700718
35127 Padova
Italy
Date: Thu, 19 Feb 1998 23:58:49 -0500 (EST)
From: Paul Gendron <gendron@dlws15.bc.edu>
Subject: #18 Answer: Realtime fast wavelet transform (WD 7.2 #14)
Answer to question 14.
There are 2 ways ( that I know of ) to solve the realtime problem
1) use border filters, start with the paper by Zixiang Xiong
Flexible Tree Structured Signal Expansions Using Time Varying
Wavelet Packets IEEE Trans. VOL 45 NO2 feb 1997
Many helpful references in this paper see [14] and [17]
2) use overlapping of tiles at borders
see my web page (paper presented at Duke International Workshop
on wavelets/statistics week) where I use this approach for
wavelet packet best basis search and soft threshold
denoiseing of seismic signals
http://ece.wpi.edu/~gendron
basically it is what 'overlap-add' are to convolution.
Paul Gendron
Date: Tue, 17 Feb 1998 19:45:01 +0000 (IST)
From: SENTHIL KUMAR T <tsenthil@giasmda.vsnl.net.in>
Subject: #19 Question: parametrization of wavelets
Dear Wavelet Digest Team,
I am an Undergraduate student working on the application of
Wavelets for ECG compression and chracterization. I am trying to
employ Genetic Algorithms for choosing an optimal wavelet from a space
of wavelets. I would like to know if there is any parametrization of
wavelet space which could be used to generate wavelets. It would help
in the application of Genetic procedure to choose the wavelet to be
used. Thanking you.
Senthil
T. SENTHIL KUMAR,
E-mail : tsenthil@giasmda.vsnl.net.in
Phone : +91-44-2364533
[Answer WD 7.5#24]
Date: Wed, 18 Feb 1998 02:54:17
From: VLADIMIR SIDORENKO <sidorenko-vladimir@usa.net>
Subject: #20 Question: Wavelet analysis of medical audio signals
Dear Waveletters
I`m looking for information and people for wavelet analysis of Medicine signals (especially audio - respiratory, cardiac and etc.)
Who interested in this field, has some information or need it - email
me. Short about me: 24, post graduate student from Ukraine, Zhitomir,
Interests: DSP, Wavelets, Programming B-).
Date: Thu, 19 Feb 1998 13:22:30 -0500
From: Sankar Basu <sankar@watson.ibm.com>
Subject: #21 Question: IIT looking for wavelet material
A few years ago, a new IIT was established in Gauhati, Assam, India.
Our colleagues there would very much appreciate our assiatance in
building up their research library on wavelets. Therefore, any
material including journals, preprints, reprints, books relating to
wavelets and their applications would be much appreciated by them.
Please send whatever you can to
The Director
IIT Gauhati
Institution of Engineers
Panbazar,
Gauhati 781001
Assam, India
Thanks very much.
Charles Miccheli
Date: Wed, 25 Feb 1998 11:46:46 +0300
From: tap@meteo.ru
Subject: #22 Question: SAS wavelet technique
I am interested in possible wavelet techniques using SAS System. Has
anybody experience in wavelet technique in SAS?
Sterin Alexandre,
tap@meteo.ru
Date: Wed, 25 Feb 1998 17:29:46 -0500
From: Nic <iaallc@travel-net.com>
Subject: #23 Question: Looking for papesr on Discrete Wavelet Multi-Tone (DWMT)
Dear Webmaster,
I would like to receive the latest information on the subject of
Discrete Wavelet Multitones. I seem to notice only paper (copy of wh
ich I already possess) on this subject.
I would appreciate if you could email me any more publications on this
subjects, or information regarding them. My email is
<ntress@site@uottawa.ca>
Thank you.
Sincerely,
Dr. N. Tressler.
[Answer WD 7.4 #14]
Date: Wed, 4 Mar 1998 12:55:36 -0600
From: Marcin Jakubowski <mjakubow@students.wisc.edu>
Subject: #24 Question: Tracking phase information
Fellows,
Is there a way to track phase information at a selected range of frequencies?
I am a second year grad student in physics at U. Wisconsin, Madison,
on turbulence in fusion tokamaks. We have a spectroscopy experiment
at General Atomics in San Diego, where we measure light fluctuations
coming from the core of the tokamak. We measure 32 distinct
locations. We are interested in following the time evolution of phase
between light fluctuations in two channels, as that provides
information on localized electric field fluctuations within the fusion
plasma. Noone has yet measured electric field fluctuations in the
core of tokamaks. Please give me some guidance on this topic. It
sounds like wavelets would be appropriate for analysis, as they are
good at following temporal behavior. Please give me some feedback.
Sincerely,
Marcin J.
Date: Sun, 15 Mar 1998 16:53:12 +0100
From: Fabio Marchioni <f.marchioni@flashnet.it>
Subject: #25 Questions: Wavelets and analysing EMG signals
I am a graduating student in Electronic Engineering working on a
degree thesis on "The performances of the amplitude estimators for
nonstationary signals". The main application is towards the amplitude
estimation and the evaluation of the EMG pattern recognition. I am
also interested in using wavelets for analising EMG signals in dynamic
conditions about the estimation of the amplitude parameter on the
pattern discrimination of the EMG envelope and relative denoising. I
would like to know where I can find complete algorithms and source
programs written in "C" for this application. Thanks in advance for
every signalling of papers, sites, books, informations and suggestions
dealing those subjects.
Thank you,
Fabio Marchioni, Rome, Italy
f.marchioni@flashnet.it
Date: Mon, 16 Mar 1998 00:14:08 -0500
From: Hugh Casement <hughcasement@compuserve.com>
Subject: #26 Question: Reinvented wavelets?
Dear Waveleteers / Chers Ondeletteurs / Liebe Wellchen-Leutchen,
Forgive a stupid question from a physicist who left the academic world 25=
years ago.
I recently tried the following experiment. I took a 2x2 block of pixels
with intensities
-------
| a | b |
|---|---|
| c | d |
-------
and applied the transformation
A = 0.5 (a + b + c + d)
B = 0.5 (a + b - c - d)
C = 0.5 (a - b + c - d)
D = 0.5 (a - b - c + d)
The inverse transformation is precisely the same, swapping upper and
lowe r case. I repeated this for all 2x2 blocks in an image, creating
four new arrays each a quarter the size. The B, C, and D arrays
contained many zeros, so would be amenable to run-length encoding. I
repeated the process on the A (sum) array, and then on its sum array,
and so on. Theoretically one cou ld take a 1024 x 768 array and
reduce it progressively to 4 x 3, though I suspect it's more
reasonable to break off the process rather sooner.
What I'd like to know is: have I reinvented wavelets? The process
seems to have little in common with the abstruse mathematics appearing
in articles on the subject in learned journals. Looking at it from a
musician's poin t of view one could say that, whereas a Fourier
transform reduces a wavefor m to its partials (arithmetic progression
of frequencies), my method reduce s it to octaves (geometric
progression). Assuming this is a kind of primitive wavelet transform,
perhaps someone could suggest a more refined method that gives better
image compression without being complicated to implement.
Hugh Casement <hughcasement@compuserve.com>
Note from the editor: You have rediscovered Haar wavelets. For a good
on-line tutorial on how to go beyond Haar look at Amara's web page:
http://www.amara.com/current/wavelet.html
<webmaster@wavelet.org>