[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 8 (August 24, 1998)
Today's Editor: Wim Sweldens
Lucent Technologies, Bell Laboratories
wim@bell-labs.com
Today's Topics:
1: Preprint: Publications not recognized on the net or literature.
2: Preprint: Wavelets and Their Associated Operators (A. Ron)
3: Preprint: Computer vision and image analysis (Y.P-Wang)
4: Preprint: Low Frequency Prioritized Coding Algorithms
5: Preprint: New Preprints and Tech Reports Available (Taswell)
6: Thesis: Construction of Splines and Wavelets on the Sphere...
7: Thesis: A new and practical generalization of Fourier analysis
8: Thesis: Interpolation Schemes and Wavelets
9: Software: Image coder using biorth. wavelets in Java
10: Meeting: Wavelets and Applications this Fall in the Swiss Alps
11: Meeting: Workshop and Seminars on Wavelets in Tokyo
12: Job: Postdoctoral Position at UVEG-Valencia (SPAIN)
13: Job: PhD student position at EPFL in Switserland
14: Contents: Special issue of the Journal of Mathematical Physics
15: Contents: J. Approx. Th. - jul-aug98 TOC
16: Contents: Numerical Algorithms 17-3,4
17: Question: WSQ bin width computation for deep-bit images
18: Question: Wavelets and clinical diagnosis
19: Question: Fourier series like calculation
20: Question: Edge detection software
21: Question: Daubechies wavelets on the interval
22: Question: History of Wavelets
23: Question: Differencial equations and wavelet
24: Question: Choosing one wavelet
25: Question: Wavelets and CDMA
26: Question: How to the decomposition tree level for wavelet packets
Current number of members: 12152
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:
Sep 7-11: Multivariate Approximation and Interpolation, Israel
Sep 28-30: Applied Computational Electromagnetics, Penn State
*Sep 28-Oct2: Wavelets and Applications this Fall, Swiss Alps
Oct 4-9: School on Wavelets in the geosciences, Delft, Netherl.
Oct 5-8: 10th GAMM-Workshop on Multigrid Methods, Bonn, Germany
Oct 27-28: Wavelets in Biology and Medicine, Hong Kong
*Oct 28-29: Workshop and Seminars on Wavelets in Tokyo, Japan
Feb 24-25: Royal Society London, UK
Feb 24-26: IEEE Information Theory Workshop, Santa Fe, NM
---
Mar 22-26: II Symposium on Artificial Intelligence, Havana, Cuba
Jul 1-7: Curves and Surfaces, St-Malo, France
Date: Wed, 05 Aug 1998 18:14:10 -0500
From: "James B. Ramsey" <ramseyj@fasecon.econ.nyu.edu>
Subject: #1 Preprint: Publications not recognized on the net or literature.
Some wavelet publications seemingly not recognized on the net.
I am including some published articles that do not seem to be
recognized in the wavelets literature.
Ramsey, James, B., George Zaslavsky, and Daniel Usikov, "An Analysis
of U.S. Stock Price Behavior Using Wavelets", Fractals, 3, No. 2,
1995, 377-389.
Ramsey, James B., and Zhifeng Zhang, " The Application of Waveform
Dictionaries to Stock Market Index Data" in Predictablility of
Complex Dynamical Systems, edt. James Kadtke and Yurii A. Kravtsov,
Springer-Verlag, 1996.
Ramsey, James B. and Zhifeng Zhang, " The Analysis of Foreign
Exchange Rates Using Wveform Dictionaries", Journal of Empirical
Finance, 4, 1997, 341-372.
Ramsey, James B. and Camille Lampart, "The Decomposition of Economic
Relationships by Time Scale Using Wavelets: Money and Income",
Macroeconomic Dynamics, vol. 2, 1998, 49-71.
Ramsey, James B. and Camille Lampart, "The Decomposition of Economic
Relationships by Time Scale Using Wavelets: Expenditure and Income,"
Studies in Nonlinear Dynamics and Econometrics, Vol. 3, No. 4, 1998,
23-42.
Thank you
James B. Ramsey
Dept.of Economics
New York Univ.
269 Mercer St.
New York, NY 10003
Office: 212-998-8947
Fax: 212-995-3932
Ramseyj@Fasecon.Econ.NYU.edu
Date: Fri, 7 Aug 1998 18:23:08 +0800 (SST)
From: "Amos Ron" <amos@math.nus.edu.sg>
Subject: #2 Preprint: Wavelets and Their Associated Operators (A. Ron)
A survey (+ a conjecture) on wavelets and refinable functions
``Wavelets and Their Associated Operators'', by Amos Ron, July 1998 (35 pp.)
It is my entry to the Texas IX meeting (i.e, The 9th International
Conference on Appproximation Theory, Nashville, 1/98), and
is actually the first survey that I have ever written.
A .ps file is at
ftp://ftp.cs.wisc.edu/Approx/texas9.ps
or
ftp://ftp.cs.wisc.edu/Approx/texas9.ps.Z
can also be obtained by going to my page
http://www.cs.wisc.edu/~amos
and using the `read me' link.
CONTENT: it is made of two disjoint parts.
the first part overviews the theory of wavelet frames: fiberization of
the shift-invariant system, quasi-wavelet systems, affine products,
the fiberization of the analysis operator of the wavelet system. it
then presents the two extension principles for constructing framelets,
and discusses in detail the tight B-spline framelets. it also shows
how the new theory applies to other topics such as band limited
wavelets and oversampling (as well as Gabor systems)
the second part is devoted to the role of the cascade/transfer
operator in the study of refinable functions. it addresses properties
like the stability of the shifts of a refinable function, the
convergence of the cascade and subdivision algorithms, and the
smoothness of the refinable function. it is written from scratch as a
short monograph, aiming at demonstrating the simplicity and the
elegance of the theory. Highlights in this part are the discrete
cascade operator, the characterization of the smoothness using that
operator, and the accelerations techniques for the cascade iterations
(`the cascade always converges, and fast': one must keep in mind that
Approximation Theory resents results of the form `the cascade operator
converges', unless one establishes the rate of convergence...)
The article ends with a conjecture concerning the convergence of the
cascade iterations. When reading the conjecture, keep in mind that I
have a (simple) proof for the one dimensional case.
Amos Ron, 8/98
Date: Sat, 08 Aug 1998 15:30:27 +0800
From: Wang Yu-Ping <wyp@haar.math.nus.edu.sg>
Subject: #3 Preprint: Computer vision and image analysis (Y.P-Wang)
1. Title: Scale-space derived from B-splines,
to be published in IEEE Trans. Pattern Analysis and
Machine Intelligence
Authors: Yu-Ping Wang, S. L. Lee
Abstract: It is well-known that linear scale-space theory in computer
vision is mainly based on the Gaussian kernel. The purpose of the
paper is to propose a scale-space theory based on B-spline
kernels. Our aim is twofold. On one hand, we present a general
framework and show how B-splines provide a flexible tool to design
various scale-space representations: continuous scale-space, dyadic
scale-space frame, and compact scale-space representation. In
particular, we focus on the design of continuous scale-space and
dyadic scale-space frame representation. A general algorithm is
presented for fast implementation of continuous scale-space at
rational scales. In the dyadic case, efficient frame algorithms are
derived using B-spline techniques to analyze the geometry of an
image. Moreover, the image can be synthesized from its multiscale
local partial derivatives. Also the relationship between several
scale-space approaches is explored. In particular, the evolution of
wavelet theory from traditional scale-space filtering can be well
understood in terms of B-splines. On the other hand, the behavior of
edge models, the properties of completeness, causality, and other
properties in such a scale-space representation are examined in the
framework of B-splines. It is shown that, besides the good properties
inherited from the Gaussian kernel, the B-spline derived scale-space
exhibits many advantages for modeling visual mechanism with regard to
the efficiency, compactness, orientation feature and parallel
structure.
Index terms: Image modeling, B-spline, wavelet, scale-space, scaling
theorem, fingerprint theorem.
2. Title: Image representations using multiscale differential operators
Author: Yu-Ping Wang
Abstract: Differential operators have been widely used for multiscale
geometric descriptions of images. The efficient computation of these
differential operators is always desirable. Moreover, it has not been
clear whether such representations are invertible. In this paper, such
problems are studied. We consider multiscale differential
representations of images using different types of operators such as
the directional derivative operators and Laplacian operators. In
particular, we provide a general approach to represent images by their
multiscale and multi-directional derivative components. For practical
implementation, efficient pyramid-like algorithms are derived using
spline technique for both the decomposition and reconstruction of
images. It is shown that using these representations various
meaningful geometric information of images can be extracted at
multiple scales; therefore, these representations can be used for edge
based image processing purposes. Furthermore, the intrinsic
relationships of the proposed representations with the compact wavelet
models, and some classical multiscale approaches are also elucidated
in the paper.
Keywords: Computer vision, Image representation, Edge detection,
B-splines, Wavelet, Scale-space
The above preprints are available at the URL:
http://wavelets.math.nus.edu.sg/~wyp/download_papers/Preprints.html
Thanks for your comments and suggestions.
WANG Yu-Ping Email: wyp@wavelets.math.nus.edu.sg
or matwyp@leonis.nus.edu.sg
Dept. of Math. Fax: (65)7795452
National University of Singapore Phone:(65)8742773(lab)
10 Kent Ridge Crescent (65)8745005(office)
Singapore 119260
http://wavelets.math.nus.edu.sg/~wyp
Date: Mon, 17 Aug 1998 09:25:50 +0800
From: zhl <zhl@ns.glc.cn.net>
Subject: #4 Preprint: Low Frequency Prioritized Coding Algorithms
Preprint: Low Frequency Prioritized Coding (LFPC)Algorithm
Author:Li Xuejun
Email:li_xj@hotmail.com
Abstract:
An improved image compression method of JPEG is presented in this
paper. The major shortage of DCT based JPEG algorithms is which can
only remove the correlation within each block and leave the
correlations between blocks untouched. LFPC can remove both
correlations and makes the quality of restored imge much better than
that of JPEG's.
Date: Sun, 23 Aug 1998 21:10:18 -0700
From: Carl Taswell <taswell@toolsmiths.com>
Subject: #5 Preprint: New Preprints and Tech Reports Available (Taswell)
The following new preprints and tech reports by Carl Taswell are available
for download from
http://www.toolsmiths.com/papers.shtml
Disjoint Sets of Daubechies Polynomial Roots for
Generating Wavelet Filters with Extremal Properties, Aug'98
Wavelet Transform Compression of Functional
Magnetic Resonance Image Sequences, Aug'98
The Systematized Collection of Daubechies Wavelets, Jun'98
Correction for the Sherlock-Monro Algorithm for
Generating the Space of Real Orthonormal Wavelets, Jun'98
Date: Fri, 24 Jul 1998 11:59:24 +0100
From: "JOCHEN GoeTTELMANN" <jochen@mat.mathematik.uni-mainz.de>
Subject: #6 Thesis: Construction of Splines and Wavelets on the Sphere...
Dear Waveleteers,
very recently I finished my PhD thesis entitled
''Construction of Splines and Wavelets on the Sphere and
Numerical Solutions to the Shallow Water Equations of Global
Atmospheric Dynamics''.
A copy can be obtained by anonymous ftp at
ftp://ftp.mathematik.uni-mainz.de/pub/papers/misc/Goettelmann/
Please use ''compress -d PhD-Thesis.ps'' to decompress.
(Warning: due to a lot of figures the original postscript file has
approx. 20 MB).
Best regards, Jochen Goettelmann
Dr. Jochen Goettelmann Department of Mathematics
Johannes Gutenberg University
Staudinger Weg 9
Phone: ++49 6131 393617 55099 Mainz
Fax: ++49 6131 394389 Germany
Date: Mon, 27 Jul 1998 15:36:02 +0800 (HKT)
From: Yuchuan Wei <weiyuch@bltda.com.bta.net.cn>
Subject: #7 Thesis: A new and practical generalization of Fourier analysis
Title: Easily-generated Function Analysis
Author: Yuchuan Wei
Yuchuan Wei was born on September 1,1966 in China. He was awarded
Bachelar of Science at Physics Dept in Lanzhou U. in 1988, Master of
Engineering at Institute of Electronics in Chinese Academy in 1994,
and Doctor of Engineering at School of Applied Science in Beijing
U. of Sci. and Tech. in 1998. He is good at electronics, mathematics,
and physics. Now he wishes to be a visiting scholar or post-doctor in
the field of signal processing, applied mathematics, mathematical
physics in a university or institue. <weiyuch@bltda.com.bta.net.cn>
Abstract of the thesis:
In this thesis we introduce easily-generated function analysis
systematically, which is a new and practical generalization of Fourier
analysis based on sine-cosine functions.
After sine-cosine function, sawtooth wave, square wave, triangular
wave and trapezoidal wave become new easily-generated functions in
modern electronics. The following questions arise naturally: 1. Can a
signal be considered as a superposition of easily- generated functions
with different frequencies? (2) How to decompose a signal into a
series of easily-generated functions with different frequencies ? (3)
How to approximate a signal best with finite easily-generated
functions? We call this basic and important problem easily generated
function analysis.
Since easily-generated function analysis has a close relation with
number theory, Chapter 1 introduces some basic concepts and
propositions in number theory, which will be used in the following
chapters.
Since square wave is the most important easily-generated function in
electronics, we first study square wave analysis. We call the function
system of square waves with different frequencies square wave system,
which is linearly independent, nonorthogonal and complete in
L2[-pi,pi]. The biorthogonal functions and orthonormalized functions
of square waves are given explicitly. Since square wave system is
complete, any signal can be approximated by a linear combination of
finite square waves to an arbitray mean-square error. By means of the
biorthogonal functions, a signal can be decomposed into a square wave
series. With the help of the orthonormalized functions, a signal can
be approximated best by a linear combination of given finite square
waves. [Ref 1]
Triangular wave and trapezoidal wave are common easily-generated
functions as well, so we next study triangular wave analysis and
trapezoidal wave analysis. Triangular wave and trapezoidal wave are
similar to sine-cosine function not only in waveform but aslo in
analysis properties. A continous function of period 2pi can be
approximated uniformly by a linear combination of triangular waves or
trapezoidal waves with different frequencies. Triangular wave system,
I- trapezoidal wave system, and II- trapezoidal wave system are bases
of L2[-pi,pi]. Therefore any signal can be synthesized by or
decomposed into a series of triangular waves or trapezoidal waves.
Additionally, the close relations of triangular wave analysis,
trapezoidal wave analysis, as well as Fourier analysis are given
explicitly in matrix notation. [Ref 2]
Easily-generated functions may have various waveforms, so we finally
study frequency analysis based on general periodic functions. In
Chapter 4 we introduce the concepts of frequency system and frequency
series based on a general periodic function. A frequency system may be
an orthogonal basis, an unconditional basis, a complete system, or an
incomplete system in the odd function subspace of L2[-pi,pi]. We
discuss what properties a periodic function needs to possess so that
its frequency system is a complete system or an unconditional basis.
For practical conveniece almost every easily-generated function in
electronics is considered as examples. A number of complete frequency
systems or unconditional bases of practical importance are given
explicitly. [Ref 3]
This makes it possible to represent a signal by easily-generated
functions with different frequencies. The results in this thesis form
a theory foundation for the technique of easily-generated function
analysis in electronics. Also this is another important application of
number theory.
Chapter 1. ABC of Number Theory
Chapter 2. Square Wave Analysis
Chapter 3. Trangular wave analysis and Trapezoidal Wave Analysis
Chapter 4. Frequency Analysis Based on General Periodic Functions
[Ref.1] Square Wave Analysis, 39(1998) No.8, J. Math. Phys.
[Ref.2] Triangular Wave Analysis and Trapezoidal Wave Analysis.(submitted)
[Ref.3] Frequency Analysis based on general periodic functions.(submitted)
17 Figures, 120 pages, in English.
Further work and cooperations:
As a new and practical generalization of classical Fourier Analysis,
Easily-generated Function Analysis is a new branch of morden harmonic
analysis. As an interdisciplinary foundamental research, it has a
close relation with electronics, signal processing, functional
analysis, number theory, physics and so on. My research is nothing
than a beginning and of course there are a lot of work to do.
For a functional analysis, the unsolved questions are: What are the
necessary and sufficient conditions for the function system
1, X(x),Y(x),X(2x),Y(2x),...,X(nx),Y(nx),...
to be (1) a complete system; (2) a basis; (3) a unconditional basis in
L2[-pi,pi]? Here X(x) and Y(x) are an even and odd function of period
2pi. These are the basic questions in frequency analysis based on
general periodic functions. A more concrete and practical question is:
For the even square wave Xs(x)=pi/4sgn(cos x) and the odd square wave
Ys(x)=pi/4sgn(sin x), is the square wave system
1, Xs(x),Ys(x),Xs(2x),Ys(2x),...,Xs(nx),Ys(nx),...
a basis of L2[-pi,pi] ? [ Ref.1 ]
Here sgn is the sign function, i.e., sgn(x)=1 for x>1; sgn(x)=0 for x=0;
sgn(x)=-1 for x<0.
For a professor of number theory, the unsolved question is:
Does the series
sin x = Ys(x)- 1/3Ys(3x)-1/5Ys(5x)+...+u(2n-1)/(2n-1)Ys[(2n-1)x] +...
converge at every point? Here u denotes the Moebius
function. [ Ref.1 ]
For an engineer of electronics, a series of schematic diagram can
become circuits and products. Many of them can become patents.
For a signal processor, easily-generated function analysis is a new
theory and idea, and it will bring you new methods. For a software
engineer a lot of formulas can become new and practical programs.
In one word, this is a new branch of morden harmonic analysis. No
mater whether you are a professor in mathematics or physics, or an
engineer in electronics or signal processor, only if you are
interested in Easily- generated Function Analysis, I would like to
coorperate with you in all possible forms including: meeting,
workshop, report, talk, training, course, publishing book, sircuit
designing, software developement and so on.
Yours sincerely,
Dr. Yuchuan Wei
Easily-generated Function Analysis Group (EGFAG),
Applied Physics Department,
Beijing University of Science and Technology,
Beijing 100083, China.
weiyuch@bltda.com.bta.net.cn
Date: Thu, 13 Aug 1998 18:11:31 -0400
From: Daniel Lemire <daniel.lemire@tintin.net>
Subject: #8 Thesis: Interpolation Schemes and Wavelets
The following wavelet-related thesis is now available as a postscript
file.
TITLE : SCH=C9MAS D'INTERPOLATION ET ONDELETTES
(INTERPOLATION SCHEMES AND WAVELETS)
=C9cole Polytechnique de Montr=E9al and Universit=E9 de Montr=E9al
Montr=E9al, Qu=E9bec
AUTHOR : D. Lemire
JURY : A. Fortin, G. Deslauriers, S.Dubuc, R. Vaillancourt and J.-M.
Lina
DATE : March 1998
LANGUAGE : French
ENGLISH ABSTRACT :
In the early 80's, the convergence of research in engineering (signal
processing), mathematics (harmonic analysis), and physics (quantum
mechanics) gave birth to wavelet theory. At the same time, Gilles
Deslauriers and Serge Dubuc (and others) came up with iterative
interpolation. It is only later, in the early 90's, when wavelet
theory became a mature subject, that the link between iterative
interpolation and wavelets became obvious. Now, many researchers in
Quebec, Alberta, France, U.S.A., Israel, Scandinavia and, Singapore
are working on iterative interpolation for its applications in
Computer Science (graphics) and in engineering
(modelisation). However, few researchers have worked on building
wavelets from iterative interpolation schemes. This is what we wanted
to do.
On the one hand, we have discovered and shown that some well-known
wavelets, the Cohen--Daubechies--Feauveau biorthogonal wavelets, are
in fact, the derivatives of a certain family of functions (called
fundamental) obtained by the iterative interpolation scheme. This same
family of functions was adapted to the interval by Mongeau in the
early 90's. Using these facts, we adapted these wavelets to the
interval. We can then handle bounded regions of the real line.
Iterative interpolation can be easily casted into a multidimensional
context. It was therefore natural to try to build multidimensional
wavelets from some iterative interpolation schemes and, in order to
apply them, it was necessary to be able to work, for example, on
rectangular regions in the plane. We have done this in the last part
of this thesis. We have included an example of an image analysis by
means of these new wavelets.
Available for download with other documents as a zip file containing a
Postscript file at :
http://www.technopole.le-village.com/lemire/publications.html
http://cafe.rapidus.net/danilemi/publications.html
Some of the boundary adapted filters derived in this thesis are available
as a text file (with explanations in English) at
http://www.technopole.le-village.com/lemire/filtres.txt
http://cafe.rapidus.net/danilemi/filtres.txt
Finally, Java implementations of most of the ideas of this thesis and
of boundary some other boundary-handling wavelets
(including Daubechies-Meyer wavelets) with applications
to signal processing (ECG, medical data,...) and numerical analysis
(O.D.E., matrix representation of operators,...) are available at
http://www.technopole.le-village.com/lemire/java_contributions.html
http://cafe.rapidus.net/danilemi/java_contributions.html
(Page is in French, but English documentation is being written.)
Thanks!
Daniel Lemire
Daniel.Lemire@Tintin.net
Daniel Lemire
http://cafe.rapidus.net/danilemi/
Date: Tue, 28 Jul 1998 13:42:16 +0800
From: Balmelli <balmelli@de.epfl.ch>
Subject: #9 Software: Image coder using biorth. wavelets in Java
Wavelet Expansion of Image Using Biorthogonal Filter Bank
Laurent Balmelli (balmelli@de.epfl.ch)
Laboratory for Audio-Visual Communications
Swiss Federal Institute of Technology
1015 Lausanne, Switzerland
1. Introduction
------------
The applet implements an iterated biorthogonal filter bank,
where four filters are involved (namely H0, H1, G0, G1) for image
coding. The leftmost image always represents the original data set
(which can be selected among a few ones using the left-bottom
pull-down), the middle one represents the coefficients expansion,
whereas the rightmost image contains the reconstructed data. The
coefficients are arranged as usually described in the literature (with
a flip!) [1]:
-----------
LL HL
-----------
LH HH
-----------
LL : low-pass on image line, low-pass on image columns
HL : high-pass " " ", low-pass " " "
...
A set of biorthogonal filters is preset (first pull-down menu), Haar
(this one is orthogonal!), linear spline and quadratic spline, with
the delay induced by the filters, since one has linear phase. A
special option, named CUSTOM, allows the user to enter its own
filters. To do so, one must edit the filter fields above and below the
diagrams depicting the analysis part and the synthesis part of the
bank, and the filter bank total delay (rightmost field), AND FINALLY
SELECT THE CUSTOM OPTION of the pull-down menu, in order to initialize
the bank with the new filters. The message board acknowledge the bank
new setting with the message:
"Filter bank has been reset and is ready for analysis."
Filtering operations are described in the next section.
2. Filtering and Selection of Reconstruction Subbands
--------------------------------------------------
Each analysis step yields four subbands (see 1.). The "analysis"
button allows to iterate the filtering on the low-pass subband and
generate the coefficients of the wavelet transform of the image. A
counter between the "analysis" and "synthesis" buttons shows the
current number of steps achieved. Symmetrically, the "synthesis"
button allows performing a single reconstruction step. When doing so,
one can choose which subbands must be taken into account during THIS
SINGLE RECONSTRUCTION STEP, by selecting or disabling the checkboxes.
The unselected subbands will be put to zero. This feature allows to
fully reconstruct the image (the image is fully reconstructed when the
counter reaches value 0) with arbitrary subbands.
3. Transform Coefficients Representation
------------------------------------
"Good" filter banks concentrates the signal energy in the low-pass
subband. For this reason, high-pass coefficients are usually of small
non-integer magnitude. Moreover, high-pass coefficients may have
arbitrary sign. Image pixel are formed of three colors (Red, Green,
Blue), each encoded using 8 bits. Simply visualizing high-frequency
components as is (e.g. generating a coefficients image by converting
the values into 8-bit integers) yields a dark, poor quality
representation. A simple method is to translate and scale the
coefficients to transfer them to a more visible space. This operation
is performed to generate the coefficients image (middle image).
However, the reconstructed image is left unchanged, unless the user
select the "enhance" checkbox. This option is particularly useful when
one wants the reconstruct an image only using the high-pass subbands
(disable the LL checkbox during reconstruction), in order to visualize
its high-frequencies (details).
4. Conclusion
----------
The applet is a didactic tool to understand iterated filter banks,
wavelet transform, biorthogonal filters and multiresolution
decomposition of images. It is part of an iterative DSP course [2],
which is currently in preparation. It is also a simple tool to test
biorthogonal filters and compare their performance.
5. References
----------
[1] Martin Vetterli and Jelena Kovacevic - Wavelets and Subband
Coding, Prentice Hall, Englewood Cliffs, NJ 07632
ISBN 0-13-097080-8
[2] Laurent Balmelli and Serge Ayer - The Interactive DSP Book,
http://lcavwww.epfl.ch/DSPBook Swiss Federal Institute of
Technology, Switzerland.
Date: Thu, 30 Jul 1998 11:22:49 -0400
From: Akram Aldroubi <aldroubi@math.vanderbilt.edu>
Subject: #10 Meeting: Wavelets and Applications this Fall in the Swiss Alps
Dear Waveletters,
There are a few places left in a workshop on Wavelets and Applications
this Fall in the Swiss Alps. The programm is made up of invited speakers
(see list below) tutorials, as well as poster sessions.
The registration fee (inc. room and board) is 650CHF.
If you are interested to participate (with or
without a poster) please send email ASAP to Jocelyne Plantefol,
secretary to the workshop (Jocelyne.Plantefol@lcav.de.epfl.ch to register).
You can also check the web site:
http://lcavwww.epfl.ch/events/WAW_call_for_papers.html
Best regards Martin Vetterli and Michael Unser
Wavelets and Applications Workshop,
Monte Verita, Ascona, Ticino, Switzerland,
Sept. 28 - Oct. 2 1998
Organizing Committee
Martin Buhmann, Mathematics, University of Dortmund
Markus Gross, Computer Sciences, ETHZ
Michael Unser, Microengineering, EPFL
Martin Vetterli, Communications, EPFL
List of Confirmed Invited Speakers:
Edward H. Adelson
Pyramids, wavelets, and the search for effective image representations
Akram Aldroubi
The Wavelet Transform and Its Application to Medicine, Biology and
Biomedical Imaging
Peter J. Burt
A Multi-Resolution Framework for Early Vision
Albert Cohen
Wavelet thresholding and functional minimization
Rafi Coifman
Orthogonal Waveform libraries as a tool for feature extraction and
characterization
Ingrid Daubechies
From nonlinear approximation to rate distortion bounds
Tony DeRose
Wavelets in Computer Graphics and Geometric Modeling
Patrick Flandrin
Reallocation for time-frequency analysis
Stephane Mallat
Deconvolution in Anti-Wavelet Bases
Yves Meyer
Wavelets and multifractal signal processing
Mark J. T. Smith
Wavelets for Motion Analysis
Gilbert Strang
The Discrete Cosine Transform
(and Limits of Refinable Functions)
Wim Sweldens
Theory and Applications of Second Generation Wavelets
Date: Sun, 09 Aug 1998 17:16:53 +0900
From: mei@trl.ibm.co.jp
Subject: #11 Meeting: Workshop and Seminars on Wavelets in Tokyo
Subject 1: Workshop on Wavelets and Wavelet-based Technologies
Date: Oct. 29-30, 1998
Place: Tokyo, Japan
see map on http://www.iwakimu.ac.jp/~susumLab/kakenwave.html
Organizers (in alphabetical order):
Kobayashi, Mei mei@trl.ibm.co.jp (English)
Sakakibara, Susumu susumu@iwakimu.ac.jp (Japanese)
Yamada, Michio yamada@ms.u-tokyo.ac.jp (Japanese)
All talks (except those by by Leland Jameson and Akram Aldroubi)
will be in Japanese. There is no registration fee, however, attendees
are asked to sign in on the participation sheet to ensure accurate
headcount for funding for future workshops. Further details will
be available on either
http://www.iwakimu.ac.jp/~susumLab/
or http://lettuce.ms.u-tokyo.ac.jp
Pls. address any questions to the organizers in the preferred
language indicated above.
Subject 2: Half-Day Seminars on Wavelets
Information on a series of half-day seminars on wavelets, held
3 to 4 times per year, which emphasize applications useful to
mechnical engineeers is available on the same server at:
http://www.iwakimu.ac.jp/~susumLab/atswave.html
Organizers:
Inoue, Hirotsugu inoueh@mep.titech.ac.jp (Japanese)
Sakakibara, Susumu susumu@iwakimu.ac.jp (Japanese & English)
The seminars are supported, in part, by the Japan Society of
Mechanical Engineers (JSME). There is no formal participation
fee, however, regular attendees are asked to pay a few thousand
yen per year to offset mailing and photcopying costs, if
possible (receipts can be issued).
The pages mentioned above are in Japanese. They can be viewed by
setting Netscape browser options for: language, font and document
encoding to kanji/Japanese.
Date: Fri, 31 Jul 1998 10:31:30 +0100 (WET DST)
From: Francesc Arandiga Llaudes <arandiga@godella.matapl.uv.es>
Subject: #12 Job: Postdoctoral Position at UVEG-Valencia (SPAIN)
POST-DOCTORAL POSITION AT UVEG - SPAIN
Departament of Matematica Aplicada.
Universitat de Valencia.
Burjassot. Valencia (SPAIN)
TMR NETWORK
Wavelets and Multiscale Methods in Numerical Analysis and Simulation.
The University of Valencia located in Burjassot-Valencia (SPAIN)
expects to offer a post-doctoral grant in the field of Wavelets and
Multiscale Methods in Numerical Analysis and Simulation and related
topics for the academic year 1998-99. This position will be available
from 1998-10-01 for 11 months.
Young scientists whose research interests are in the area of partial
differential equations, numerical analysis and/or Scientific
computation are encouraged to apply.
The following topics: Discretization of Non Linear Problems, Fast
Computation of Non Linear Terms, Applications to Conservation Laws,
Wavelets, Multiresolution Techniques, and related areas of studies are
among the interests of our group.
Candidates must be nationals of a European Community Member State
other than Spain or a State associated with the TMR Program and must
not be over 35 years of age. People who are interested should submit
their applications (Curriculum Vitae, a list of published works,
research interest and two letters of recommendation) before
1998-09-15. Correspondence should be sent to:
TMR Postdoc Programme - Wavelets in numerical simulation.
Attn: Rosa Donat
Departament of Matematica Aplicada.
Universitat de Valencia.
C/ Dr. Moliner 50, 46100 Burjassot.
Valencia (SPAIN)
tel 34-96 386 47 27,
fax 34 96 386 40 85,
e-mail donat@uv.es
Paco Arandiga |e-mail:
Departament de Matematica Aplicada |arandiga@godella.matapl.uv.es
Facultad de Matematicas |arandiga@uv.es
C/ Dr. Moliner n. 50 |Phone (346) 386-47-27
46100 Burjassot (Valencia) |Fax (346) 386-47-35
SPAIN |Fax (346) 386-40-85
Date: Thu, 13 Aug 1998 11:11:37 +0200
From: Andrzej Drygajlo <andrzej.drygajlo@lts.de.epfl.ch>
Subject: #13 Job: PhD student position at EPFL in Switserland
Ph.D. STUDENT POSITION for the Signal Processing Laboratory (LTS) of
the Swiss Federal Institute of Technology Lausanne (EPFL), Switzerland
The LTS EPFL Speech Processing Group offers a Ph.D. student position
for research in robust speaker recognition. This 3-year position is
funded by the Swiss National Science Foundation.
The work consists of adapting Hidden Markov Model (HMM) based
text-dependent and Gaussian Mixture Model (GMM) based text-independent
speaker verification in securized interactive man-machine voice
communication systems to real acoustical and environmental conditions.
More specifically, the emphasis of the research will be placed on the
combination of multiresolution wavelet packet analysis and the new
approaches in feature extraction and stochastic model compensation.
This position must be filled as soon as possible. The candidate, with
a university degree, should be able to work independently and to
interact with the existing research team. Candidates are expected to
have good knowledge of digital signal processing, particularly in the
domain of speech processing, and have a sound experience in
programming. Because this position involves regular contacts in the
framework of European projects, a working knowledge of English and
French is required.
Interested candidates are invited to send a CV as soon as possible to:
Dr Andrzej Drygajlo, EPFL-DE/LTS, CH-1015 Lausanne, Switzerland, Email:
Andrzej.Drygajlo@epfl.ch, Tel.: +41 21 693 43 28, Fax: +41 21 693 46 60
Kind regards,
Dr. Andrzej Drygajlo
Date: Thu, 6 Aug 98 16:47:16 +0100
From: Bruno Torresani <torresan@cptsx1.univ-mrs.fr>
Subject: #14 Contents: Special issue of the Journal of Mathematical Physics
Dear colleagues
The Journal of Mathematical Physics publishes in its August issue
a special volume
"Wavelets and Time-Frequency Analysis"
Please find below the table of contents of the special volume.
Yours sincerely
Bruno Torresani
Centre de Physique Theorique
CNRS-Luminy, case 907,
13288 Marseille Cedex 09, France
Bruno.Torresani@sophia.inria.fr
B. Torresani:
Introduction
S. T. Ali:
A general theorem on square-integrability: vector coherent states
P. Aniello, G. Cassinelli, E. De Vito & A. Levrero:
Wavelet transforms and discrete frames associated to semidirect products
H. Fuehr:
Continuous Wavelet Transforms with Abelian Dilation Groups
J.P. Antoine & P. Vandergheynst:
Wavelets on the n-sphere and related manifolds
T. Paul & A. Uribe:
Weighted trace formula near an hyperbolic trajectory and complex orbits
P. Flandrin:
On separability, Positivity and minimum Uncertainty in
Time-Frequency Energy Distributions
G. Matz & F. Hlawatsch:
Time-Frequency Transfer Function Calculus of Time-Varying Systems
Based on a Generalized Underspread Theory.
J. Bertrand & P. Bertrand:
Symbolic calculus on the time-frequency half-plane
P. Frick, A. Grossmann & P. Tchamitchian:
Wavelet analysis of signals with gaps
P.G. Lemarie-Rieusset:
Some remarks on the Navier-Stokes equations in R3
J. Lewalle:
Formal Improvements on the Solution of the Wavelert-Transformed
Poisson and Diffusion Equations
S. Jaffard:
Oscillation Spaces: Properties and Applications to Fractal and
Multifractal Functions
A. Arneodo, E. Bacry & J.F. Muzy:
Random cascades on wavelet dyadic trees
C.A. Guerin & M. Holschneider:
Time-dependent scattering on fractal measures
J. Zak:
Orthonormal Sets of Localized Functions for a Landau Level
J.P. Gazeau, J. Patera & E. Pelantova:
tau wavelets on the plane
G. Bernuau:
Wavelets basis adapted to a self-similar quasicrystal
Y. Wei & N. Chen:
Square wave analysis
Date: Tue, 11 Aug 1998 17:29:36 -0400 (EDT)
From: Thomas Hogan <hogan@math.ohio-state.edu>
Subject: #15 Contents: J. Approx. Th. - jul-aug98 TOC
Table of Contents: J. Approx. Theory, Volume 94, Number 1, July 1998
Paul Erd\H{o}s (1913--1996)
1--41
Vadim I. Filippov
On the completeness and other properties of some function systems in
$L_p$, $0<p<\infty$
42--53
F. S. De Blasi and J. Myjak
On a generalized best approximation problem
54--72
Alexei Zhedanov
On some classes of polynomials orthogonal on arcs of the unit circle
connected with symmetric orthogonal polynomials on an interval
73--106
F. Marcell\'an and I. A. Rocha
Complex path integral representation for semiclassical linear
functionals
107--127
R. W. Barnard, G. Dahlquist, K. Pearce, L. Reichel, and K. C. Richards
Gram polynomials and the Kummer function
128--143
F. J. Mu\~noz-Delgado, V. Ram\'\=ED{r}ez-Gonz\'alez, and D. C\'ardenas-Mo=
rales
Qualitative Korovkin-type results on conservative approximation
144--159
Shunsheng Guo, Cuixiang Li, Yiguo Sun, Ge Yang, and Shujie Yue
Pointwise estimate for Szasz-type operators
160--171
Table of Contents: J. Approx. Theory, Volume 94, Number 2, August 1998
Luca Brandolini
Imbedding theorems for Lipschitz spaces generated by the weak-$L^p$
metric
173--190
M. A. Skopina
Local convergence of Fourier series with respect to periodized wavelets
191--202
Jean-Paul Penot
Proximal mappings
203--221
Douglas Bridges and Wang Yuchuan
Constructing best approximations on a Jordan curve
222--234
S. B. Damelin
The Lebesgue function and Lebesgue constant of Lagrange interpolation
for Erd\H{o}s weights
235--262
S. Fridli and F. Schipp
Strong approximation via Sidon type inequalities
263--284
Sen-Yen Shaw
Non-optimal rates of ergodic limits and approximate solutions
285--299
Vilmos Totik
Counterexample to a problem on tensor product approximation
300--305
Yasuo Kageyama
Generalization of the left Bernstein quasi-interpolants
306--329
Date: Fri, 24 Jul 1998 00:29:35 +0200
From: Baltzer Science <mailer@ns.baltzer.nl>
Subject: #16 Contents: Numerical Algorithms 17-3,4
Numerical Algorithms 17 (1998) 3,4
J.C. Butcher
ARK methods up to order five 193-221
Herbert H.H. Homeier
On the stability of the J transformation 223-239
J.W. Schmidt and M. Bastian-Walther
Algorithm for constructing range restricted histosplines 241-260
Vishwani D. Agrawal, David Lee and Henryk Wozniakowski
Numerical computation of characteristic polynomials of Boolean functions
and its applications 261-278
Christian Rey and Franck Risler
A Rayleigh--Ritz preconditioner for the iterative solution to large scale
nonlinear problems 279-311
Miodrag S. Petkovic, Dorde Herceg and Snezana Ilic
Safe convergence of simultaneous methods for polynomial zeros 313-331
N. Mansouri and J.P Kernévez
Estimation of parameters in nonlinear problems 333-343
G. Papageorgiou, I.Th. Famelis and Ch. Tsitouras
A P-stable singly diagonally implicit Runge--Kutta--Nyström method 345-353
Erratum 355-355
Book reviews 357-358
Date: Thu, 23 Jul 1998 17:53:57 +0900
From: Ken Sutherland <ken@j-mac.co.jp>
Subject: #17 Question: WSQ bin width computation for deep-bit images
Howdy!
I have attempted to code the bin width computation published in the
FBI digitized fingerprint compression (WSQ) standard. It seems to work
reasonably well for 8-bit images, but I'd like to use the same
procedure for "deep-bit" 12 and 16-bit (signed and unsigned) gray
scale medical images. However, the computed bin widths are
unpredictable, resulting in over (or under) compression.
I believe that the unpredictable results are being cause by huge
variance values computed from the subbands. The variance values are
huge especially with signed 16-bit images because the orginal pixel
values can range between plus and minus 32,768.
I realize this is quite a long shot, but has anyone out there
attempted to adapt the interative procedure for computing bin widths
to deep-bit images? Or is the algorithm applicable strictly to 8-bit
images?
References to the WSQ standard (.ps files) can be found here:
http://www.wavelet.org/wavelet/digest_02/digest_02.18.html#9
Thanks!
Ken Sutherland
J-MAC SYSTEM INC.
ken@j-mac.co.jp
http://www.j-mac.co.jp/
http://www.voicenet.co.jp/~ken
Date: Tue, 28 Jul 1998 15:42:16 +0800
From: Yipeng Tang <typ@public1.sta.net.cn>
Subject: #18 Question: Wavelets and clinical diagnosis
Question: Where can I obtain some reading materials about wavelet in
signal processing and clinical diagnosis? If you have the answer,
please write to me via both email and letter. And I am available at
typ@public1.sta.net.cn. Because I cannot access to Internet now,
please mail me your materials including books or thesis to the
following address:
Yipeng Tang
Room 501,
266 Xi Kang Road
Shanghai 200040
P.R.China
Thanks,
Tang
Date: Wed, 29 Jul 1998 09:20:56 +0000
From: krzem@kki.net.pl
Subject: #19 Question: Fourier series like calculation
Dear Wavelet Team
I am a mathematician. I made the algorithm for numerical analysis
sums of functions:
f(x) =A1*cos(P1+F1*x)+A2*cos(P2+F2*x)+...+ An*cos(Pn+Fn*x). / 1 /
/not Fourier series ! / and test program for two components.
Program can calculate amplitude, phase and frequency for two
components. Sequence data can included a part of function , smaller
than cycles of each components. Minimal quantity of numbers requir
for depend on noise level and exactitude of numbers. Redundancy
in this method eliminates noise and digital-effect.
Algorithm can perform : Ak-amplitude,Pk -phase, Fk -frequency for any
components. Algorithm has mathematical base different than Fourier
Transformation , matching pursuit or estimation. From Princip
algorithm is dedicate only for sums / 1 /
Effect is reached by calculation middle value of constant which
definite sums of cosine (not amplitude, phase, frequency).
I would verify and compare this algorithm with similar algorithm or
program. Question and Problem : If algorithm is interest, perform
program for any (k) quantity components ?
Thanks in advance.
Aleksander Adam Krzemien
Zbicka 8/23
32-065 KRZESZOWICE
POLAND
(012)2820613
krzem@kki.net.pl
Date: Sun, 02 Aug 1998 17:54:31 +0800
From: niuysbit <niuysbit@163.net>
Subject: #20 Question: Edge detection software
Dear Wavelet Digest,
Sorry for interrupting you, I want to
apply wavelet in image registration.
Is there any free software and it's scource code
for image edge(contour) detection using local
modulus maxima(Canny' method) and zero crossing
(Marr-H. method, Laplace and Gauss method)?
Please give me a reply. Thank you!
Yours, Niu Yongsheng
niuysbit@163.net
Date: Mon, 03 Aug 1998 09:34:40 +0200
From: Armelle DAULLE <daulle@labs.polycnrs-gre.fr>
Subject: #21 Question: Daubechies wavelets on the interval
Question
I am involved in some research using Daubechies wavelets on the
interval [0;1] Could anyone give me some clues as to constructing the
basis on the interval ? I mean how do I construct the wavelets near
zero and near one ? I was told there was a special linear combination
of inner wavelets to do, but I do not know enough on the subject.
Thanks in advance
Armelle DAULLE <daulle@labs.polycnrs-gre.fr>
Date: Sun, 2 Aug 1998 17:27:32 +0200
From: umb@marlowe.dtro.e-technik.tu-darmstadt.de (Uwe Meyer-Baese)
Subject: #22 Question: History of Wavelets
Question: History of Wavelets/Error in the Matlab Toolbox documentation?
In the MatLab toolbox manual on page 1-28 "History of Wavelets" you
find the statement: 'The first recorded mention of the term "wavelet"
was in 1909, in a thesis by Alfred Haar.' Thats not true. I have a
copy of the thesis. The function family is called "orthogonal function
decomposition", but not wavelet, small wave, or "kleine Welle" (the
german term), or ondelette (the French term).
In fact there are two questions:
1. When did the word wavelet (or ondelette, or kleine Welle) first
appear in the literature ?
2. When did it first appear in the meaning of a wavelet function as we
know it now ?
The answer to 1 may be Huygens, the physician (ca. 1678), who used the
term to describe "small wave packets" but I have no proof of this
statement which I have heard once at a wavelet conference. Regarding
the answer 2 we find in the Vetterli/Kovacevic book, that it should
come from "ondelette" in French which was introduced by Morlet in the
early 80's.
Uwe Meyer-Baese
EMAIL: uwe@alpha.ee.ufl.edu
URL: http://www.dtro.e-technik.th-darmstadt.de/umb/umb.html
Date: Tue, 4 Aug 1998 14:18:49 -0300 (EST)
From: "Nubia S.D. Brito" <nubia@petri.dee.ufpb.br>
Subject: #23 Question: Differencial equations and wavelet
Hi,
I would like to get some information and references on solving the
linear differential equations by using of wavelet transform. Any help
would be appreciated Thanks in advance.
Nubia Brito
email: nubia@dee.ufpb.br
Date: Thu, 6 Aug 1998 15:40:21 -0200 (GRNLNDDT)
From: Fernanda Marques <fernanda@gpeb.ufsc.br>
Subject: #24 Question: Choosing one wavelet
Dear Waveleters,
Theory in wavelets is very hard, isn't it? That's because I am not
sure that I had understood everything I read. So, I will appreciate
very much if someone could tell me, if there exists something already
developed in therms of "how to choose best wavelet?" to be applied in
a certain work? I am trying to cross-correlate one kind of pattern
with the several types of wavelets, but interpretation of results has
became more difficult than I had tought that it could be.
Please, help me!
With best regards,
Fernanda Isabel Marques, Electrical Eng., M.Sc.
Biomedical Engineering Research Group - GPEB
Federal University of Santa Catarina - Brazil
Date: Mon, 17 Aug 1998 18:34:29 -0300 (EST)
From: Marcos Marcelo Lopes <mcelopes@e-net.com.br>
Subject: #25 Question: Wavelets and CDMA
I am a student of Electrical Engineering at the Federal University
of Bahia (UFBA), in Brazil.
I was asked to make an undergraduate level research about the usage
of the Wavelet Transform in Communication Systems, more specificly in CDMA
systems.
I'd appreciate if you could help me on this subject, even indicating
people or web sites related.
Sincerely,
Marcos Marcelo Lopes
[Answers WD 7.9 #14]
Date: Sun, 23 Aug 1998 15:52:38 +1000
From: Dennis Lee <dennis.lee@eng.monash.edu.au>
Subject: #26 Question: How to the decomposition tree level for wavelet packets
Dear editor
Application of wavelet packets to fault diagnosis for feature extraction
Many examples using wavelet packets are chosen of decompostion tree
level 3 or 4 for signal analysis. Is there any mathematcal theory or
other references to explain the proper procedure to correctly select
decomposition tree level of wavelet packets ?
For example, I used wavelet packet and decomposition tree level 3 and
got good result to analyse a signal for limited amount of data. Level
3 is selected although the maximum tree level of my signals may be 10.
Hence, why to select level 3 but not level 4? I hope to find any
papers or guideline to correctly select the best decomposition tree
level for particular signals.
Lee Dennis
Monash University, Australia
dennis.lee@eng.monash.edu.au
[Answer WD 7.11 #15]
[Answer WD 7.12 #21]
<webmaster@wavelet.org>