[Home]
[Back Issues]
[Search]
[Vol 1]
[Vol 2]
[Vol 3]
[Vol 4]
[Vol 5]
The Wavelet Digest - Volume 3, Issue 9 (June 11, 1994)
Today's Editor: Wim Sweldens
sweldens@math.sc.edu
Today's Topics:
1. Question: Wavelets for preprocessing acoustic information
2. Question: measuring the roughness of surfaces using wavelets
3. Question: Wavelets and fractals
4. Preprint: On wavelets and applications from Flinders University
5. Question: Use of wavelets in noise removal
6. Question: looking for software on wavelets and pde's
7. Preprint: Fast wavelet transforms for matrices arising from BEMs.
8. Preprint: Wavelet-Based Multifractal Analysis
9. Question: A proper grid for locating orthogonal wavelets ...
10. Question: Wavelets and time series
11. Announce: First IEEE international conference on image processing
12. Contents: JAT Vol. 77, No.1, April 1994
Submissions for Wavelet Digest:
E-mail to wavelet@math.sc.edu with "submit" as subject.
Subscriptions for Wavelet Digest:
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.
Archive site, preprints, references and back issues:
Anonymous ftp to ftp.math.sc.edu (129.252.12.3),
directory /pub/wavelet.
Gopher server: gopher.math.sc.edu.
Current number of subscribers: 3435
Calender of events:
June 24-27: Workshop course on Wavelets Filter Banks, Boston WD 3.3 #3
June 29-30: Tutorials and workshop on wavelets, Singapore WD 3.6 #8
July 11-15: Wavelet Sessions 14TH IMACS World Congress, Atlanta WD 3.6 #4
July 27-29: SPIE: Wavelet Applications, San Diego WD 3.2 #9
Aug 17-20: Wavelets for Engineering Applications, College Station WD 3.7 #12
Sep 25-28: Visual Communications and Image Processing, Chicago WD 2.16 #3
Sep 26-30: Total Positivity and its Applications, Jaca, Spain WD 3.2 #8
Oct 25-28: IEEE Symposium on Time-Frequency Analysis, Philadelph. WD 3.6 #11
Nov 1-2: Wavelets in Medicine and Biology, Baltimore, Maryland WD 2.17 #11
Nov 13-16: Artificial Neural Networks in Engineering, St. Louis WD 3.4 #8
>Nov 13-16: IEEE international conference on image processing WD 3.9 #11
Nov 16-18: Meeting on wavelets and statitics, Grenoble, France WD 3.5 #3
---
Jan 8-12: Approximation Theory, College Station WD 3.7 #4
From: A.Dennison <ADEN@mechman.mm.swin.edu.au> Swinburne University
Subject: Question: Wavelets for preprocessing acoustic information
I am working on my masters project in which I am using acoustic
information for non-destructive evaluation of weld quality. From
papers on oceanic signals (similar signal envelope to my data) I
understand that the Wavelet transform is a good method for
extracting feature vectors for subsequent analysis. The article
"Wavelets and Signal Processing" (IEEE SP Magazine 10/91) also
suggested that the phase of the wavelet transform contained more
information than the amplitude.
I would appreciate any information on using Wavelets for
preprocessing non-stationary signals; techniques for analysing the
information contained in the transformed data and comments on using
complex wavelet transforms - as every inplementation I have found
only calculates the magnitude.
Thank-you for any help.
Andrew Dennison
The CIM Centre Melbourne, AUSTRALIA
From: uhl@edvz.sbg.ac.at
Subject: Question: measuring the roughness of surfaces using wavelets
Dear wavelet-friends,
I'm interested in measuring the roughness of surfaces using
wavelet-methods.
Has there already been done some work on this topic ? Could you please
give me some hints (references,...) ?
e-mail: uhl@edvz.sbg.ac.at
Thanks in advance,
Andreas UHL e-mail: uhl@edvz.sbg.ac.at
Research Inst. for Softwaretechnology Salzburg University, Austria
phone : ##43/662/8044/5308 fax : ##43/662/8044/6727
Have a look at the WWW: http://www.mat.sbg.ac.at/people/uhl.html
From: Julia Schnabel <J.Schnabel@cs.ucl.ac.uk>
Subject: Question: Wavelets and fractals
Dear Waveletters,
I am presently trying to estimate the fractal dimension of medical
images via wavelet transforms (using Numerical Recipes' routines and
the procedure proposed by Stephane Mallat in "A Theory for
Multiresolution Signal Decomposition: The Wavelet Representation",
IEEE PAMI-11, No.5, 7/89). Since this approximation appears to be by
far too coarse and computationally expensive when applied in 2D in
local neighbourhoods for every image pixel, I am wondering whether
anybody has some experiences or helpful suggestions.
Thank you in advance,
Julia Schnabel
J.Schnabel@cs.ucl.ac.uk
University College London Dept. of Computer Science
From: J. Kautsky <jarka@ist.flinders.edu.au>
Subject: preprints: On wavelets and applications from Flinders University
Preprints on Wavelets and Applications available at Flinders University
Following ftp area contains preprints from a group based at:
Flinders University, School of Information Sciences and Technology,
Discipline of Mathematics
Mailing address:
Dr. Jaroslav Kautsky Phone : (international)-61-8-201 2360
Mathematics, IST
Flinders University FAX : (international)-61-8-201 2904
GPO Box 2100
ADELAIDE SA 5001
Australia email : jarka@ist.flinders.edu.au
or na.kautsky@na-net.ornl.gov
Anonymous ftp address: ftp.cs.flinders.edu.au directory /pub/wavelets
(N.B. from some sites, address adam.ist.flinders.edu.au has to be used)
Contents: (in reversed order - newest first)
No: filename:
6) bio1.dvi bio1.ps -- (submitted to Linear Algebra and Applications)
J. Kautsky & R. Turcajova: Discrete biorthogonal wavelet transforms as
block circulant matrices
5) reg1.dvi reg1.ps -- (to appear in the Proc. of
"Software and Alg's in Num. Math")
J. Kautsky & R. Turcajova: Extremal characterization of regular wavelet
matrices and an algorithm for their evaluation.
4) symm.dvi symm.ps -- (to appear in Numerical Algorithms)
R. Turcajova: Factorizations and Construction of Linear Phase Paraunitary
Filter Banks and Higher Multiplicity Wavelets
3) shift -- (to appear in Numerical Algorithms)
R. Turcajova & J. Kautsky: Shift products and factorizations of wavelet
matrices
2) jkrt2 -- (to appear in Linear Algebra and Applications)
J. Kautsky & R. Turcajova: Pollen Product Factorization and Construction
of Higher Multiplicity Wavelets.
1) jk1.dvi jk1.ps -- (appeared in Applications of Mathematics,
38 (1993), 169-193 )
J. Kautsky: AN ALGEBRAIC CONSTRUCTION OF DISCRETE WAVELET TRANFORMS
0) ftpinfo this info and abstracts
From: karlsenr@cc.tacom.army.mil
Subject: Question: Use of wavelets in noise removal
I am currently investigating the use of wavelets in the extraction of
a known signal from other signals plus noise. I have a few questions
regarding this:
1) Does it make sense to try and build a wavelet specific to the known
signal or should one use a standard wavelet and compare the wavelet
transforms of the known signal and the signal under investigation?
In other words, does it help if the wavelet looks like the signal
that one is trying to find?
2) Is there any software available that already does this and/or what
are some good papers that address this application?
3) In note 5.12 on page 166 of Daubechies' 'Ten Lectures on Wavelets'
book, she says that if one knows only the samples f(n) then in order
to perform the subband filtering (pg 156) one should first compute
the coefficients <f,phi0,n>. Is this necessary? I have been applying
the subband filtering directly to the f(n).
Thanks,
Rob Karlsen
From: Trappe <trappe@msmailgw1.arlut.utexas.edu>
Subject: Question: looking for software on wavelets and pde's
Dear Colleagues,
I am working on solving partial differential equations using wavelets and am
interested in finding out if there is any software in existance (Matlab,
Mathematica, or C code preferrable) that deals with this problem.
Any leads to finding some code would be appreciated.
Sincerely,
Wade Trappe
Applied Research Labs
The University of Texas at Austin
P.O. Box 8029
Austin, Tx 78713-8029
Email: trappe@msmailgw1.arlut.utexas.edu trappe@math.utexas.edu
From: dbond@cs.cornell.edu (David M. Bond)
Subject: Preprint: Fast wavelet transforms for matrices arising from BEMs.
The following preprint is available through an anonymous ftp to
"ftp.tc.cornell.edu" in the directory "pub/tech.reports". The file
name is TR174.ps (The form is uncompressed PostScript). A list of all
available online CTC technical reports is located in the file
"tr.list", and the corresponding abstracts are located in the file
"tr.abstracts" in the same directory.
Cornell Theory Center Technical Report CTC94TR174 :
FAST WAVELET TRANSFORMS FOR MATRICES
ARISING FROM BOUNDARY ELEMENT METHODS
Dave M. Bond, Stephen A. Vavasis
Cornell Theory Center
Cornell University
Ithaca, New York, 14853
For many boundary element methods applied to Laplace's equation in
two dimensions, the resulting integral equation has both an integral
with a logarithmic kernel and an integral with a discontinuous kernel.
If standard collocation methods are used to discretize the integral
equation we are left with two dense matrices. We consider expressing
these matrices in terms of wavelet bases with compact support via a
fast wavelet transform as in Beylkin, Coifman and Rokhlin. Upper
bounds on the size of the wavelet transform elements are obtained.
These bounds are then used to show that if the original matrices are
of size $N \times N$, the resulting transformed matrices are sparse,
having only $O(N \log N)$ significant entries. Some numerical results
will also be presented.
Unlike Beylkin, Coifman and Rokhlin who use the fast wavelet
transform as a numerical approximation to a continuous operator
already expressed in a full wavelet basis of $\L2(\R)$, we think of
the fast wavelet transform as a change of basis matrix for a finite
dimension, and apply it to a discretized function or matrix. As a
result, we can use this fast wavelet transform as a ``black box''
transformation in existing boundary element codes.
Sincerely, Dave Bond
719 Engineering & Theory Center Cornell University Ithaca, N.Y. 14853
e-mail : dbond@cs.cornell.edu, phone : (607) - 254 - 8872
From: Anthony Davis <davis@climate.gsfc.nasa.gov>
Subject: Preprint: Wavelet-Based Multifractal Analysis
Dear Waveleteers,
I have noticed in past issues of the WD many questions about time-series
analysis and some about multifractal approaches. In the following paper we
survey statistical scale-invariance, mostly in the analysis mode, from a
wavelet standpoint.
TITLE: Wavelet-Based Multifractal Analysis of Nonstationary and/or
Intermittent Geophysical Signals
AUTHORS: Anthony Davis, Alexander Marshak and Warren Wiscombe
ABSTRACT:
We show how some of the most popular wavelet transforms can be used to
compute simple yet dynamically meaningful statistical properties of a
one-dimensional dataset representative of a geophysical field or
time-series. The observed properties of turbulent velocity fields and of
liquid water density in clouds are used throughout as examples along with
theoretical models drawn from the literature. We introduce a diagnostic
tool called the "mean multifractal plane," simpler than (but representative
of) the two well-known multifractal hierarchies of exponents H(p) and D(q),
associated with structure functions and singularity analysis respectively.
It is used to demonstrate the pressing need for a new class of stochastic
models having both additive/nonstationary and multiplicative/intermittent
features.
The paper will published as a chapter of "Wavelets in Geophysics" (Eds. Efi
Foufoula-Georgiou and Praveen Kumar, 365 pp, Academic Press) to appear in
July 1994. For a preprint, please contact me at:
Anthony Davis,
NASA Goddard Space Flight Center,
Climate & Radiation Branch (code 913),
Greenbelt, MD 20771
E-mail: davis@climate.gsfc.nasa.gov
Tel: (310) 731 9300; Fax: (310) 731 1180
From akbar@ce.su.oz.au Wed Jun 8 05:56:55 1994
From: Ali Akbar Safavi <akbar@ce.su.oz.au>
Subject: Question: A proper grid for locating orthogonal wavelets ...
Question: A proper grid for locating orthogonal wavelets of different
scales in a multi-resolution approximation :
Dear Wavelet-ers
I follow Mallat's work on " A Theory for Multi-Resolution Signal Representation:
The Wavelet Representation "
and publishied in : IEEE Trans. on pattern Analysis and... (1989).
I have a question and appreciate your comments.
My Question is:
If we construct a grid (whose interval in proportional to 2^(-j))
for locating wavelets at different scales,
then, should the wavelets at one scale be orthogonal to all the
wavelets of the other scales, in orther to have a multi-resolution
analysis without redundant information in our detail (the difference
between approximation at two different scales)?
(assumptions:
1. the orthogonality of translation of wavelets for each
particular scale is satisfied;
2. we only look at orthogonal wavelets;
3. the wavelet may be centered at t=0 or t=1/2; )
A negative answer accept the two following grids, while a positive answer accept
only the first grid,
(First) finest res. * * * *
* * *
* *
coarsest res. *
----------------------------------------------------------
(Second) finest res. * * * * *
* * *
coarsest res. * *
The grid that Daubechies uses in page 10-11 of "ten lectures on wavelets"
to give an example of multi-resolution with Haar wavelet is the same as
the first grid. Moreover, in that example she proves orthogonality of Haar
wavelets of all different scales.
Best regards
akbar safavi
From: jim@gatsibm.larc.nasa.gov
Subject: Question: Wavelets and time series
Hey gang!
I need some basic information concerning the variety of wavelet transforms
that are commonly used. I want to analyze time series and I'm having
trouble interpreting my results.
I have seen two or three methods for generating wavelet coefficients:
1. Continuous wavelet transform
2. Discrete wavelet transform
3. Multiresolution Analysis
In (1), we basically convolve our function (signal) with a wavelet, say
for example the Mexican Hat function. We can specify the scaling and
translation parameters to generate a mapping from $\Re \mapsto \Re^2$.
From there one can look at the phase and modulus and try to locate
features that may be of interest, such as singularities.
In (2) we are essentially performing the same steps as in (1), again
generating the mapping.
Given these coefficients, I have seen methods to generate wavelet
syntheses of our function [Percival] as well as power spectrums [Hudgins].
Now in (3), one typically generates coefficients by {\it S. Mallat's}
pyramidal decomposition scheme.
I understand the mechanics of all the above but the interpretation of
results eludes me. For example, how do the coefficients in (1) and (2)
relate to those generated in (3)? Once I have generated and displayed
$W(a,b)$ how do we relate scale to frequency. I'm aware of the
inverse relationship between scale and frequency but relating a particular
scale number with frequency (like we typically see in Fourier Power
Spectra).
Any basic help with the above is welcome. One thing to note is that I
have read most of 'popular' literature on wavelets, thus my understanding
of the mechanics. I have yet to find something equitable dealing with
interpretation. Perhaps I'm not seeing the trees!?
References:
Percival, D.. 'An Introduction fo Spectral Analysis and Wavelets'
Hudgins, L.. 'Wavelet Transforms and Spectral Estimation'
Jawerth, B., Sweldens, W.. 'An Overview of Wavelet Transforms and
Multiresolution Analysis'
Thanks in Advance and Apologies for the length!
Jim Craft
jim@gatsibm.larc.nasa.gov
From: icip@pine.ece.utexas.edu (International Conf on Image Processing Mail Box)
Subject: Announce: First IEEE international conference on image processing
FIRST IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING
November 13-16, 1994
Austin Convention Center, Austin, Texas, USA
Sponsored by the Institute of Electrical and Electronics
Engineers (IEEE) Signal Processing Society, ICIP-94 is the
inaugural international conference on theoretical, experimental
and applied image processing. It will provide a centralized,
high-quality forum for presentation of technological advances and
research results by scientists and engineers working in Image
Processing and associated disciplines such as multimedia and
video technology, as well as image processing applications in
areas such as the biomedical sciences and geosciences.
ICIP-94 will be held in the recently completed state-of-the-art
Convention Center in downtown Austin. The Convention Center is
situated two blocks from the Town Lake, and is only 12 minutes
from Robert Meuller Airport. It is surrounded by many modern
hotels that provide comfortable accommodation for $75-$125 per
night.
An exciting range of EXHIBITS, TUTORIALS, SPECIAL PRODUCT
SESSIONS, and SOCIAL EVENTS will be offered.
Tutorials: Sunday, Nov. 13, 1994
Session 1 AM Bernd Girod - Compression of Still and Moving Images
Session 2 AM Robert Haralick - Mathematical Morphology
Session 3 PM Martin Vetterli/Jelena Kovacevic
Wavelets and Subband Coding
Session 4 PM Richard Blahut - Maximum-Likelihood Image Synthesis:
An Introduction to Imaging Algorithms
For further details about ICIP-94, please contact:
Conference Management Services
3024 Thousand Oaks Drive
Austin, Texas 78746
Tel: 512/327/4012; Fax:512/327/8132
or email: icip@pine.ece.utexas.edu
From: Marilyn Radcliff <radcliff@math.ohio-state.edu>
Subject: JAT Vol. 77, No.1, April 1994
Journal of Approximation Theory, Volume 77, Number 1, April 1994
Table of Contents
J. M. Carnicer and W. Dahmen. Characterization of Local Strict Convexity
Preserving Interpolation Methods by $C^1$ Functions, 2-30.
I. E. Gopengauz. Pointwise Estimates of the Hermitian Interpolation,
31-41.
D. S. Lubinsky and F. Moricz. The Weighted $L_p$-Norms of Orthonormal
Polynomials for Freud Weights, 42-50.
Harold Widom. The Asymptotics of a Continuous Analogue of Orthogonal
Polynomials, 51-64.
Bernhard Beckermann, Jacek Gilewicz, and Valeri Kaliaguine. On the
Definition and Normality of a General Table of Simultaneous Pad\'e
Approximants, 65-73.
Vitaly Maiorov. Linear Widths of Function Spaces Equipped with the
Gaussian Measure, 74-88.
Avram Sidi. Rational Approximations from Power Series of Vector-Valued
Meromorphic Functions, 89-111.
<webmaster@wavelet.org>