[Home] [Back Issues] [Search] [Calendar] [Vol 1] [Vol 2] [Vol 3] [Vol 4] [Vol 5] [Vol 6] [Vol 7] [Vol 8] [Vol 9]

The Wavelet Digest, Volume 9, Issue 1 (February 8, 00)




Today's Editor: Wim Sweldens
                Lucent Technologies, Bell Laboratories
                http://wim.sweldens.com



Today's Topics:

     1: Book:     The Functional and Harmonic Analysis of Wavelets and Frames
     2: Preprint: Compression in time-domain computational electromagnetics
     3: Preprint: Restoration and mirror wavelet deconvolution (Kalifa/Mallat)
     4: Preprint: A module frame concept for Hilbert C*-modules
     5: Preprint: Analysis Of The Non-Stationary Processes
     6: Preprint: Subdivision Surfaces: A New Paradigm for Thin-Shell FEM
     7: Preprint: Curvature Smoothness of Subdivision Surfaces
     8: Preprint: A 12dB improvement in progressive geometry compression
     9: Preprint: Normal meshes
    10: Thesis:   Electromagnetic Integral Equations With Wavelets
    11: Thesis:   Wavelets and ill-posed problems
    12: Meeting:  2 Wavelet workshops
    13: Meeting:  CBMS-NSF Regional Conference 2000 in Applied Mathematics.
    14: Meeting:  Search of Venue for the FoCM 2002 Conference
    15: Meeting:  Numerical Methods in Multiphase Flow and Transport Processes
    16: Job:      Postdoctoral Research Position in Optics/Computational Math.
    17: Job:      Temporary positions at MINGLE
    18: Job:      NASA-NRC Postdoc and Senior Research Associateships
    19: Job:      Faculity positions at UCLA
    20: Job:      Internship, IBM, wavelets + times series prediction
    21: Job:      PhD position at EPFL, CH
    22: Contents: J. Approx. Th. - nov99 TOC
    23: Contents: J. Approx. Th. - nov99 TOC
    24: Contents: J. Approx. Th. - dec99 TOC
    25: Contents: Advances in Computational Mathematics 12  (2000)
    26: Contents: J. Approx. Th. - jan00 TOC
    27: Answer:   Image decompression artifacts (WD 8.10 #19)
    28: Answer:   Signal classification by wavelets (WD 8.9 #27)
    29: Answer:   Image decompression artifacts (WD 8.10 #19)
    30: Answer:   Integer based wavelets and lifting (WD 8.10 #23)
    31: Question: How make the model of currency forecast
    32: Question: Applying WT to Binary Image Data
    33: Question: Looking for wavelet software in C or C++
    34: Question: Wavelet ask for help 
    35: Question: Looking for earlier versions of WavBox.
    36: Question: request for information
     
Current number of members: 15354

E-mail:
   General help:      help@wavelet.org
   Add yourself:      add@wavelet.org
   Remove yourself:   remove@wavelet.org with Subject: $ADDRESS
   Publish in the WD: publish@wavelet.org 
   Publishing policy: policy@wavelet.org

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

Calendar of events:

*Feb 18-19: Wavelet and Multiscale Software, Aachen, Germany
*Feb 21-22: Wavelet Methods in Elastoplasticity, Aachen, Germany
 Feb 29:    Time-scale and time-frequency analysis, London.
 Apr 25-29: Scaling, Multifractals in Geophysics, Nice, France
*Mar 6-10:  Numerical Methods in Multiphase Flow, Germany
 Mar 12-16: Australian Society for Exploration Geophysics, Perth
 May 17-20: Trends in Approximation Theory 2000, Nashville, TN
*May 22-26: CBMS-NSF Regional Conference, St-Louis
 Jun 5-9:  Wavelets at the SIMAI 2000 conference, Ischia, Italy
 Jun 29 - Jul 4: 5th Curves and Surfaces, Oslo, Norway
 Jul 13-17: Steve Smale's 70th birthday, Hong Kong
 Jul 17-19: Wavelet Applications, Tokyo, Japan
 Jul 30 - Aug 2: SPIE Wavelet Applications VIII, San Diego
 Jul 30 - Aug 2: Lifting session at the SPIE meeting,  San Diego
 Jul 30 - Aug 2: IEEE Multimedia And Expo, New York










Date: Mon, 7 Feb 2000 10:07:19 +0100 (MEZ)
From: Michael Frank <frank@mathematik.uni-leipzig.de>
Subject: #1 Book: The Functional and Harmonic Analysis of Wavelets and Frames

The purpose of this message is the announcement of the content of a new
publication in the field of wavelet and frame theory:

      'The Functional and Harmonic Analysis of Wavelets and Frames'

                    Joint Mathematics Meeting 1999
  AMS Special Session on the Functional and Harmonic Analysis of Wavelets,
            January 13-14, 1999, San Antonio, Texas, USA.
                     (L. W Baggett, D. R. Larson)
                     Contemp. Math. v. 247 (2000)

A. Aldroubi, P. Basser
   'Reconstruction of vector and tensor fields from sampled discrete data'
   pp. 1-15

L. G. Baggett, K. D. Merrill
   'Abstract harmonic analysis and wavelets in R^n'
   pp. 17-27

R. Balan
   'Density and redundancy of the noncoherent Weyl-Heisenberg superframes'
   pp. 29-41

J. J. Benedetto, M. T. Leon
   'The construction of multiple dyadic minimally supported frequency wavelets
   on R^n'
   pp. 43-74

L. Brandolini, G. Garrigos, Z. Rzeszotnik, G. Weiss
   'The behaviour at the origin of a class of band-limited wavelets'
   pp. 75-91

O. Bratteli, P. E. T. Jorgensen
   'Convergence of the cascade algorithm at irregular scaling functions'
   pp. 93-130

P. G. Casazza, O. Christensen, A. J. E. M. Janssen
   'Classifying tight Weyl-Heisenberg frames'
   pp. 131-148

P. G. Casazza, D. Han, D. R. Larson
   'Frames for Banach spaces'
   pp. 149-182

J. Courter
   'Construction of dilation-d wavelets'
   pp. 183-205

M. Frank, D. R. Larson
   'A module frame concept for Hilbert C*-modules'
   pp. 207-233

J. Gasch, J. E. Gilbert
   'Triangularization of Hankel operators and the bilinear Hilbert transform'
   pp. 235-248

R. F. Gundy
   'Two remarks concerning wavelets: Cohen's criterion for low-pass filters
   and Meyer's theorem on linear independence'
   pp. 249-258

D. Han, D. R. Larson, M. Papadakis, Th. Stavropoulos
   'Multiresolution analysis of abstract Hilbert spaces and wandering
   subspaces'
   pp. 259-284

G. Strang, V. Strela, D.-X. Zhou
   'Compactly supported refinable functions with infinite masks'
   pp. 285-296

E. Weber
   'Applications of the wavelet multiplicity function'
   pp. 297-306

For more details see www.ams.org or the January 2000 issue of the Notices of
the AMS.
                                              M. Frank, Leipzig, Germany.



Date: Thu, 30 Dec 1999 08:14:32 +0200 (IST)
From: Yehuda Leviatan <leviatan@ee.technion.ac.il>
Subject: #2 Preprint: Compression in time-domain computational electromagnetics

The following preprint (in postscript form) may be obtained by sending
e-mail request to leviatan@ee.technion.ac.il

Title: Iterative Impedance Matrix Compression (IMC) for the
Time--Domain Integral Equation using Spatiotemporal
Multiresolution Analysis

Authors: Y. Shifman and Y. Leviatan

Abstract: A novel method of moments approach to the solution of
time--domain integral--equation (TDIE) formulation of electromagnetic
scattering problems is presented. The method is based on a
spatiotemporal multiresolution analysis. This analysis facilitates a
basis from which a small number of expansion functions is selected via
an iterative procedure and utilized to model the unknown current
distribution.  In contrast to marching-on-in-time (MOT) sequential
procedures, the proposed method models the unknown current
simultaneously at all the time steps within the time frame of
interest.  This new method is applied to a one--dimensional problem of
electromagnetic plane wave interaction with a dielectric slab. A
comparison of the computed results with results based on the analytic
solution demonstrates that the method is capable of attaining accurate
results while achieving substantial reduction in computation time and
resources.



Date: Thu, 13 Jan 2000 22:13:59 +0100 (CET)
From: Jerome Kalifa <Jerome.Kalifa@polytechnique.fr>
Subject: #3 Preprint: Restoration and mirror wavelet deconvolution (Kalifa/Mallat)

The first paper is application-oriented while the second paper is
math-oriented.

1- Jerome Kalifa, Stephane Mallat, Bernard Rouge
"Minimax Deconvolution in Mirror Wavelet Bases", submitted 
to IEEE Transactions on Image Processing, 1999.

Inverting a low-pass convolution of signals in presence of additive
noise is numerically unstable. To solve these ill-posed inverse
problems, we study diagonal estimators in an orthogonal basis. We
construct nearly optimal non-linear estimators in a minimax sense for
the restoration of bounded variation signals and images, with a
thresholding in a wavelet packet basis called mirror wavelet
basis. These results are applied to the deconvolution of satellite
images provided by the French spatial agency (CNES).

Postscript and pdf files of this paper can be found at

http://wave.cmap.polytechnique.fr/~kalifa/deconv.html

2- Jerome Kalifa, Stephane Mallat
"Thresholding Estimators for Inverse Problems and Deconvolutions",
submitted to Annals of Statistics, 2000.

Thresholding algorithms in an orthonormal basis are studied to
estimate noisy signals degraded by a linear operator whose inverse is
not bounded. For signals in a given set, sufficient conditions are
established on the basis to obtain a maximum risk which is
asymptotically equivalent to the minimax risk. Deconvolutions with
transfer functions having a zero at high frequencies are examples of
unstable inverse problems. A new ``mirror wavelet'' basis is
constructed to obtain a deconvolution risk which is asymptotically
equivalent to the minimax risk over bounded variation signals. This
thresholding estimator is used to restore satellite images.

Postscript and pdf files of this paper can be found at 
 
http://wave.cmap.polytechnique.fr/~kalifa/diaginverse.html

Jerome Kalifa
Center for Biomedical Engineering, MC 8904, Schapiro
Columbia University, New York,  NY 10027



Date: Fri, 4 Feb 2000 11:50:21 +0100 (MEZ)
From: Michael Frank <frank@mathematik.uni-leipzig.de>
Subject: #4 Preprint: A module frame concept for Hilbert C*-modules

Michael Frank, David R. Larson
'A module frame concept for Hilbert C*-modules'
preprint 1999, to appear in: Contemp. Math., 2000.

The goal of the present paper is a short instructive introduction to a
general module frame theory in C*-algebras and Hilbert C*-modules. The
reported investigations rely on the idea of geometric dilation to
standard Hilbert C*-modules over unital C*-algebras that possess
orthonormal Hilbert bases, and of reconstruction of the frames by
projections and other bounded module operators with suitable
ranges. We obtain frame representation and decomposition theorems, as
well as similarity and equivalence results. The relative position of
two and more frames in terms of being complementary or disjoint is
investigated in some detail. In the last section some recent results
of P. G. Casazza (in J. Fourier Anal.  Appl. 4(1999), 727-732) are
generalized to our setting. The Hilbert space situation appears as a
special case (cf. D. Han, D. R. Larson, Memoirs of the AMS, to
appear). For the details of many proofs we refer to our basic
publication 'Frames in Hilbert C*-modules and C*-algebras', preprint,
1998. One benefit for classical frame theory is the additionally won
degree of freedom treating (norm-completed) tensor products A x H ,
where A is a C*-algebra and H is a Hilbert space, without major
changes in the involved techniques.

Universitaet Leipzig, Mathematisches Institut, D--04109 Leipzig, Germany,
frank@mathematik.uni-leipzig.de.
Dept. Mathematics, Texas A&M University, College Station, TX 77843, USA,
larson@math.tamu.edu.



Date: Fri, 04 Feb 2000 17:18:18 +0300
From: Alexander Prygunov <Alexander.Prygunov@mstu.edu.ru>
Subject: #5 Preprint: Analysis Of The Non-Stationary Processes

The New Index For Current Analysis Of The Non-Stationary Processes
Alexander I. Prygunov, Murmansk State Technical University.

Abstract
The new non-dimensionally index of the oscillation's forms based on
wavelets and phase mapping is offered in the paper. Unlike of Lyapunov's
index, which is used to estimation of amplitude stability (stability by
Lyapunov), the new index can be directly use to estimation of systems=92
structurally stability at the case of non-stationary vibration. The
first examples of method's application to solving  of non-stationary
vibration's problems  are presented.



From: Peter Schroeder <ps@cs.caltech.edu>To: publish@wavelet.org
Subject: #6 Preprint: Subdivision Surfaces: A New Paradigm for Thin-Shell FEM

We would like to announce the availability of our preprint
"Subdivision Surfaces: A New Paradigm for Thin-Shell Finite-Element
Analysis" for web download at

http://multires.caltech.edu/pubs/thinshell.pdf

Subdivision Surfaces: A New Paradigm for Thin-Shell Finite-Element
Analysis, (Fehmi Cirak and Michael Ortiz and Peter Schroeder),
International Journal for Numerical Methods in Engineering, 2000 (in
print).

Abstract: We develop a new paradigm for thin-shell finite-element
analysis based on the use of subdivision surfaces for: i) describing
the geometry of the shell in its undeformed configuration, and ii)
generating smooth interpolated displacement fields possessing bounded
energy within the strict framework of the Kirchhoff-Love theory of
thin shells. The particular subdivision strategy adopted here is
Loop's scheme, with extensions such as required to account for creases
and displacement boundary conditions. The displacement fields obtained
by subdivision are H2 and, consequently, have a finite Kirchhoff-Love
energy.  The resulting finite elements contain three nodes and element
integrals are computed by a one-point quadrature. The displacement
field of the shell is interpolated from nodal displacements only.  In
particular, no nodal rotations are used in the interpolation.  The
interpolation scheme induced by subdivision is nonlocal, i.e., the
displacement field over one element depend on the nodal displacements
of the element nodes and all nodes of immediately neighboring
elements.  However, the use of subdivision surfaces ensures that all
the local displacement fields thus constructed combine conformingly to
define one single limit surface. Numerical tests, including the
Belytschko et al. obstacle course of benchmark problems, demonstrate
the high accuracy and optimal convergence of the method.  ------- End
of forwarded message -------



Date: Mon, 7 Feb 2000 13:42:26 -0800 (PST)
From: Peter Schroeder <ps@cs.caltech.edu>
Subject: #7 Preprint: Curvature Smoothness of Subdivision Surfaces

We would like to announce the availability of our preprint "Curvature
Smoothness of Subdivision Surfaces" (Caltech TR-00-03) for download at

http://multires.caltech.edu/pubs/h2.pdf

Curvature Smoothness of Subdivision Surfaces, (Ulrich Reif and Peter
Schroeder), Caltech TR-00-03 (submitted for publication).

Abstract: We examine the smoothness properties of the principal
curvatures of subdivision surfaces near extraordinary points. In
particular we give an estimate of their $L_p$ class based on the eigen
structure of the subdivision matrix. As a result we can show that the
popular Loop and Catmull-Clark schemes (among many others) have square
integrable principal curvatures justifying their use as shape
functions in FEM treatments of the thin shell equations.



Date: Tue, 8 Feb 2000 13:23:20 -0500
From: Wim Sweldens <wim@bell-labs.com>
Subject: #8 Preprint: A 12dB improvement in progressive geometry compression

Title: Progressive Geometry Compression
Authors: Andrei Khodakovsky, Peter Schroeder, Wim Sweldens 

Abstract: We propose a new progressive compression scheme for
arbitrary topology, highly detailed and densely sampled meshes arising
from geometry scanning. We observe that meshes consist of three
distinct components: geometry, parameter, and connectivity
information. The latter two do not contribute to the reduction of
error in a compression setting. Using semi-regular meshes, parameter
and connectivity information can be virtually eliminated. Coupled with
semi-regular wavelet transforms, zerotree coding, and subdivision
based reconstruction we see improvements in error by a factor four
(12dB) compared to other progressive coding schemes.

Status: Preprint. Submitted January 2000.

/who/wim/papers/compression/



Date: Tue, 8 Feb 2000 13:23:27 -0500
From: Wim Sweldens <wim@bell-labs.com>
Subject: #9 Preprint: Normal meshes

Title: Normal Meshes

Authors: Igor Guskov, Kiril Vidimce, Wim Sweldens, Peter Schröder 

Abstract: Normal meshes are new fundamental surface descriptions
inspired by differential geometry. A normal mesh is a multiresolution
mesh where each level can be written as a normal offset from a coarser
version. Hence the mesh can be stored with a single float per
vertex. We present an algorithm to approximate any surface arbitrarily
closely with a normal semi-regular mesh. Normal meshes are useful in
numerous applications such as compression, filtering, rendering,
texturing, and modeling.

Status: Preprint. Submitted January 2000.

/who/wim/papers/normalmesh/



Date: Wed, 05 Jan 2000 08:59:07 -0500
From: "Miller, Richard E" <richard.e.miller@lmco.com>
Subject: #10 Thesis: Electromagnetic Integral Equations With Wavelets

The following dissertation is available at
http://www.umi.com/hp/Products/Dissertations.html

Publication Number:  AAT 9943540

ELECTROMAGNETIC INTEGRAL EQUATIONS WITH WAVELETS

Author:                    MILLER, RICHARD EARL

School:                   TEXAS A&M UNIVERSITY 

Advisor:                   NEVELS, ROBERT

Abstract: Integral equations that are used to study electromagnetic
  problems generate dense matrices when solved by the moment
  method. The use of a wavelet multiresolution analysis basis is
  presented to allow for thinning (removing the smallest matrix
  elements) the matrix to produce a sparse form. The wavelet basis is
  implemented in the moment method matrix equation by direct expansion
  and application of the discrete wavelet transform. A comparison of
  different wavelets in the discrete wavelet transform implementation
  of a wavelet multiresolution analysis basis in the solution of the
  matrix equation is conducted. The lower/upper decomposition direct
  solution method and the iterative biconjugate gradient method are
  used to solve the matrix equations.

Richard
       Richard E. Miller, Ph.D.
       Principal Member Engineering Staff
       Lockheed Martin Naval Electronics & Surveillance Systems -
       Moorestown
       199 Borton Landing Rd
       P.O. Box 1027
       Moorestown, NJ 08057-0927
       Telephone  856-722-4076           Fax  856-722-2207
       E-mail: richard.e.miller@lmco.com



Date: Wed, 26 Jan 2000 17:43:33 +0100 (CET)
From: Christophe BERNARD <bernard@cmapx.polytechnique.fr>
Subject: #11 Thesis: Wavelets and ill-posed problems

Title: Wavelets and ill-posed problems: optic flow and scattered data
interpolation

PhD dissertation, Ecole Polytechnique, France, November 1999

Author: Christophe Bernard

Supervisor: Stephane Mallat

Abstract:

This thesis describes a way to handle two different ill-posed
problems: optic flow measurement and scattered data interpolation,
where they compete among others with regularization techniques.

In either case, the thesis describes a complete innovative numerical
technique to solve the problem, and motivates the suggested technique
with consistency or convergence results.

The technique described to solve optical flow is based on a discrete
analytic wavelet analysis of the differential optic flow equation,
with a number of technical enhancements that make the approach
competitive w.r.t to existing methods, while remaining computationally
inexpensive.

The technique described to solve the scattered data interpolation
problem is based on an adaptive wavelet tree construction on the basis
of the positions of the samples, called "allocation". It is motivated,
analysed in detail and compared thoroughly to other
methods. Convergence proofs are given, and the method is tested
numerically.

The thesis is available in French or English, postscript or PDF, from
the following web pages. It can be downloaded as a whole or
chapterwise.

French:

  http://www.cmap.polytechnique.fr/~bernard/these/

English:

  http://www.cmap.polytechnique.fr/~bernard/these/index-en.html

Christophe BERNARD
Centre de Mathematiques Appliquees, Ecole Polytechnique,
91128 Palaiseau cedex, FRANCE  tel 33-1-69334665 -- fax 33-1-69333011
http://www.cmap.polytechnique.fr/~bernard/



Date: Mon, 17 Jan 2000 09:18:01 +0100
From: Karsten Urban <urban@igpm.RWTH-Aachen.DE>
Subject: #12 Meeting: 2 Wavelet workshops

                     Announcement                      

The EC-TMR (Traveling and Mobility for Researches) network 
"Wavelets and Multiscale Methods in Numerical Analysis and Simulation"
                 organizes two workshops                         

"Wavelet and Multiscale Software" 
Aachen, Germany, February 18-19, 2000 

The aim of this workshop is to give an overview on existing software
tools for wavelets and multiscale methods. Moreover, guidelines for
documentation of and access to the software will be discussed.
Computing facilities (Silicon Graphics workstations, PCs under Linux)
will be available.

"Wavelet Methods in Elastoplasticity" 
Aachen, Germany, February 21-22, 2000   

The aim of this workshop is to review the current state of the art in
wavelet based numerical solution methods for elastoplasticity
problems.  Moreover, it will be discussed which kind of realistic
elastoplasticity problems are appropriate to be treated by wavelet
methods.

Invited speakers (confirmed):
Dietrich Braess, Bochum (Germany)
Klaus Hackl, Bochum (Germany)
Giovanni Naldi, Milano (Italy)

For both workshops:

Organizers: Wolfgang Dahmen, Karsten Urban 

Location: RWTH Aachen, 
          Institut für Geometrie und Praktische Mathematik, 
          Germany 
    
Conference fee: -- 

Registration: as soon as possible by e-Mail:       
                urban@igpm.rwth-aachen.de

Accomodation: the Aachen team will organize the local accomodation.  
        Please give precise dates of arrival and departure with
        your registration.

How to travel to Aachen:
        http://www.aachen-tourist.de/

For further questions please contact:
        http://www.igpm.rwth-aachen.de/~urban/tmr/soft-workshop.html
        http://www.igpm.rwth-aachen.de/~urban/tmr/plasto-workshop.html

Local organization: 
        Karsten Urban
        RWTH Aachen
        Inst. f. Geometrie u. Praktische Mathematik
        Templergraben 55
        52056 Aachen, Germany
        Phone : +49 / 241 / 806338
        Fax   : +49 / 241 / 8888317
        e-Mail: urban@igpm.rwth-aachen.de



Date: Thu, 20 Jan 2000 14:26:01 -0600 (CST)
From: Grant Welland <welland@welland.cs.umsl.edu>
Subject: #13 Meeting: CBMS-NSF Regional Conference 2000 in Applied Mathematics.

The University of Missouri - St. Louis is hosting a CBMS Regional Conference.
The principal speaker, title and dates of the conference are

                              David Donoho
                           Stanford University

                Interaction of Harmonic Analysis, 
           Statistical Estimation and Data Compression,

                           May 22 - 26, 2000.

Special lectures will be given in addition to the lectures of Professor Donoho.
These lectures will be given by 

        Amir Averbuch,          Tel Aviv University, 
        Emmanuel Candes,        Stanford University,
        Charles Chui,           University of Missouri - St. Louis,
        R. R. Coifman,          Yale University,
        Ingrid Daubechies,      Princeton University
        Ron Devore,             University of South Carolina,
        David Field,            Cornell University,
        Iain Johnstone,         Stanford University,
        Naoki Saito,            University of California, Davis,
        Martin Vetterli,        Swiss Federal Institute of Technology, Lausanne 
                                and the University of California, Berkeley and
        Victor Wickerhauser,    Washington University.

Registration and a welcoming cookout will occur on May 21, 2000,
at the south Residence Hall on the U. of MO - St. Louis south campus.
        
For further information visit the Conference web site at 
http://www.math.umsl.edu/CBMS. 

The conference organizer is 

                        Grant Welland
                        Mathematics and Computer Science   
                        University of Missouri - St. Louis
                        St. Louis, Missouri 63121
                        e-mail address: welland@math.umsl.edu
                        phone (314) 516 6356 
                        fax   (314) 516 5400
 


Date: Tue, 1 Feb 2000 15:51:55 -0500 (EST)
From: Ron DeVore <devore@math.sc.edu>
Subject: #14 Meeting: Search of Venue for the FoCM 2002 Conference

         Search of venue for the FoCM 2002 Conference

Following the very successful second international conference on the
Foundations of Computational Mathematics (FoCM) at Oxford University
in July 1999 which attracted 320 participants and a high-powered cast
of foremost world authorities from all branches of mathematics, the
Executive Committee of FoCM is now investigating possible venues and
calling for proposals for the next FoCM conference in Summer 2002.
Although the Committee is open-minded as to the geographic location,
there is a definite body of opinion to the effect that this is North
America's turn (after Brazil and the UK).

If you contemplate the idea of organising the next FoCM conference,
please drop an e-mail to Endre.Suli@comlab.ox.ac.uk as soon as
possible to explore this further. Formal applications (up to 2 pages
long) should be sent to him by 31 March, 2000, including details about
lecture room facilities, quality and cost of local accommodation and
subsistence, convenience of travel to the region, and potential
sources of funding.

The FoCM Executive Committee expects to announce the venue and the
likely time of the next FoCM conference in April 2000.

The aim of FoCM as an organisation is to explore and foster the
interactions and establish a common agenda between computational
mathematics, pure mathematics and computer science through
conferences, workshops and series of publications which include the
new Springer-Verlag journal ``Foundations of Computational
Mathematics'', edited by Mike Shub, and the new Cambridge University
Press monograph series ``Library of Computational Mathematics''
published under the FoCM imprint.

Further details are available from the official FoCM website:

         http://www.focm.net/



Date: Fri, 4 Feb 2000 15:40:16 +0100
From: "Listowner" <listowner@baltzer.nl>
Subject: #15 Meeting: Numerical Methods in Multiphase Flow and Transport Processes

We like to have your attention for the following meeting:

Subsurface Modeling:
Numerical Methods in Multiphase Flow and Transport Processes

March 06 - 10, 2000

International Association for Hydraulic Research;
Institut für ComputerAnwendungen im Bauingenieurwesen, TU Braunschweig;
European Engineering Graduate School of Environment Water (IAHR-EGW).

Course Directors:
Prof. R. Falta, Clemson University
Prof. R. Helmig, TU Braunschweig

Information at the following address:
www.tu-bs.de/institute/cab/weiterbildung/NM

Sincerely,
Baltzer Science Publishers



Date: Mon, 10 Jan 2000 10:28:29 -0700 (MST)
From: Curt Vogel <vogel@gauss.math.montana.edu>
Subject: #16 Job: Postdoctoral Research Position in Optics/Computational Math.

A postdoctoral research position concerning Computational Methods in
Atmospheric Optics will be available beginning Aug. 20, 2000. Research
will be carried out at Montana State University, in collaboration with
the USAF Starfire Optical Range in New Mexico and the Maui High
Performance Computing Center in Hawaii. A PhD in Computational or
Applied Mathematics, Optical Sciences, or related field is required by
start date. Experience programming in MATLAB or C++ is
essential. Applicants should have expertise in one of the following
three areas, as well as some familiarity with the other two: 1)
Computational methods for Inverse Problems; 2) Mathematical models for
the propagation of light through the atmosphere; and 3) High
performance computing. More information is available on the web at
http://www.math.montana.edu/~vogel/Postdoc. Screening will begin
Feb. 10, 2000.



Date: Tue, 25 Jan 2000 11:04:28 +0100
From: Ewald Quak <Ewald.Quak@math.sintef.no>
Subject: #17 Job: Temporary positions at MINGLE

                   Job advertisement

Several temporary positions, both doctoral and post-doctoral, are now
available within a new EU research training network called MINGLE
(Multiresolution in Geometric Modelling).  These positions will
normally be between 6 months and two years in duration and will
provide an excellent opportunity to gain research experience abroad.

The main objective of the project is to train young European
researchers, both in mathematics and computer science, in various
aspects of multiresolution in geometric modelling.  Relevant research
topics include: hierarchical decomposition of triangulations,
generation of smooth surfaces from irregular triangle meshes,
thinning, decimation, compression of meshes, remeshing, wavelets,
application of multiresolution representations to computer graphics,
and data structures and algorithms for storage and manipulation.

MINGLE will run from 2000 to 2003 and involves research teams
from the following nine partners:

  SINTEF Applied Mathematics, Norway
   Contact: Michael Floater <mif@math.sintef.no>
  Tel Aviv University, Israel
   Contact: Nira Dyn <niradyn@math.tau.ac.il>
  Munich University of Technology, Germany
   Contact: Armin Iske <iske@mathematik.tu-muenchen.de>
  Technion, Israel
   Contact: Craig Gotsman <gotsman@virtue3d.com>
  Max Planck Institute, Saarbrucken, Germany
   Contact: Hans-Peter Seidel <hpseidel@mpi-sb.mpg.de>
  Joseph Fourier University, Grenoble, France
   Contact: George-Pierre Bonneau <Georges-Pierre.Bonneau@imag.fr>
  University of Cambridge, UK
   Contact: Neil Dodgson <nad@cl.cam.ac.uk>
  University of Genova, Italy
   Contact: Leila De Floriani <deflo@disi.unige.it>
  Systems in Motion AS, Norway
   Contact: Paal-Robert Engnaes <preng@sim.no>

Further details are available on the webpage

  http://www.oslo.sintef.no/mingle/

(especially concerning the eligibility rules). If you consider making
an application, please contact the partner or partners you are
interested in. Women are especially encouraged to apply.

For general enquiries about MINGLE, please contact the
coordinating partner via Michael Floater (mif@math.sintef.no) or
Ewald Quak (ewq@math.sintef.no).



Date: Thu, 27 Jan 2000 16:31:37 -0800 (PST)
From: Marty Brenner <gonzo@reseng2.dfrc.nasa.gov>
Subject: #18 Job: NASA-NRC Postdoc and Senior Research Associateships

Contributed by: Marty Brenner 
                martin.brenner@dfrc.nasa.gov

       NASA/RRA/NRC Postdoctoral and Senior Research Awards 
              NASA Dryden Flight Research Center

Proposed Research Title:
------------------------ 
Nonlinear Analysis of Aerospace Systems using Multiresolution Methods

NASA Resident Research Associateships (RRA) are administered by the 
National Research Council (NRC). These opportunities at NASA Dryden 
are open only to US citizens.

Postdoctoral Research Associateships are awarded to those with a PhD
less than five years, and Senior Research Associateships to those with
the doctorate five years or more. They are usually for 1-2 years, but 
Senior awards for periods of three months or longer are considered. 
A Research Associate must conduct his research full-time and on-site.

Evaluations are conducted in June, October, and February by the NRC.
Application deadlines are Apr 15, Aug 15, and Jan 15, respectively. 
Current annual stipend is about $50K for Postdoc and $80K for Senior.

Problem definition:
-------------------
The ability to identify and control nonlinear phenomena is extremely 
important to the aerospace community. Understanding aeroelastic
nonlinear dynamics is becoming vital for advanced design configurations.
Stability estimation from flight data must account for nonlinearity to 
distinguish linear from nonlinear dynamics for model development.

Confidence bounds are desired in identification schemes to guarantee system 
stability within a flight regime. Realistic stability prediction requires
system identification consistent with uncertainty and disturbance models.

Application of multiscale wavelet methods is an attractive approach to 
this problem. Adaptive signal decompositions for signal processing of 
nonlinear and time-varying systems is a major focus of research at NASA
Dryden. Algorithms are needed to generate adaptive multiresolution
representations for identification in aircraft dynamics analysis.

Dryden is studying linear and nonlinear identification algorithms and 
methodologies for in-flight aeroservoelasticity (aero+structures+controls).
Intelligent data processing and analysis methods are desired for accurate 
control-oriented stability determination during flight testing.
This integrated research effort requires innovative digital signal 
processing, identification, and robust system theory applications.

Contact:                
--------
Marty Brenner
NASA Dryden Flight Research Center
MS 4840D / RS
Edwards, California 93523-0273
Phone: 661-258-3793 (starting in March: 661-276-3793)
Email: martin.brenner@dfrc.nasa.gov

See:
http://www/trc/Postdoc/nrc.html
http://www.national-academies.org/rap



Date: Fri, 28 Jan 2000 09:47:01 -0800 (PST)
From: Abeer Alwan <alwan@icsl.ucla.edu>
Subject: #19 Job: Faculity positions at UCLA

Faculty Positions in Signal Processing
Electrical Engineering Department
University of California, Los Angeles

The Electrical Engineering Department invites applications from highly 
qualified persons for tenure-track faculty positions in the School of 
Engineering and Applied Science at the Assistant Professor level. 

The department is particularly interested in candidates with
experience in Speech, Image and Biomedical Signal Processing.
Candidates with a Ph.D. in Electrical Engineering and an excellent
research and teaching record are encouraged to apply. Applications
including a curriculum vitae, names of at least 3 referees, teaching
and research statements (no longer than 1 page each), and three
selected reprints should be sent, by March 1st 2000 to:

Professor Tatsuo Itoh
Electrical Engineering Department
School of Engineering and Applied Science
66-147A Engr IV
University of California, Los Angeles
Los Angeles, CA 90095

The University of California, Los Angeles is an Equal Opportunity Employer.



Date: Sun, 30 Jan 2000 17:16:01 +0100 (MET)
From: "F.D. Murtagh" <fmurtagh@cdsxb6.u-strasbg.fr>
Subject: #20 Job: Internship, IBM, wavelets + times series prediction

Internship starting May 2000, IBM TJ Watson research center, must have
PhD and experience of wavelet/multiscale methods for signal/time
series modeling and prediction.

Contact F Murtagh, f.murtagh@qub.ac.uk



Date: Wed, 2 Feb 2000 17:28:15 +0100
From: "Gianpaolo Evangelista" <gianevan@lcavsun1.epfl.ch>
Subject: #21 Job: PhD position at EPFL, CH

Ph.D. position at EPFL
Adapted Representations for the Analysis, Synthesis and Processing of Audio
Signals

A 3-year Ph.D. position is available at the Audio-Visual Communications
Laboratory (Prof. Martin Vetterli), Department of Communication Systems,
Swiss Federal Institute of Technology, Lausanne, Switzerland, within an
audio communication project funded by the Swiss National Funds. 
The candidate will perform research in the domain of audio and musical
signal processing. 
The focus of this project is on efficient representations of acoustic
signals for feature extraction, musical instrument recognition, automated
transcription, structured audio coding and digital audio effects. 
Possible topics include physical and perceptual modeling, adaptive wavelet
transforms and atomic representations.

Prerequisites: The preferred candidates are expected to have successfully
attended a Master level program or graduate school in Electrical
Engineering, Physics, Applied Mathematics or related field. Candidates who
did not complete or did not enroll in any Master program are requested to
attend classes in the pre-doctoral school program in our Department. Info
on the school may be found at
http://dscwww.epfl.ch/EN/graduate/default.asp.  Partially earned credits
will be considered for transfer.

Position: The successful candidate will be hired full-time or 75% part-time
at the research/teaching assistant level (full-time gross salary ~60,000
CHF /year, commensurate to experience). Funding is already available and
the candidate is expected to start within this academic year or no later
than September 2000. He/she is expected to complete all the requirements
for earning the title of Ph.D. in 3 years.

Application:    Please e-mail inquiries or applications including CV and
references to 
Dr. Gianpaolo Evangelista, Gianpaolo.Evangelista@epfl.ch
DSC-LCAV, Ecublens - CH-1015 Lausanne - SWITZERLAND



Date: Fri, 21 Jan 2000 12:42:19 -0500 (EST)
From: Thomas Hogan <hogan@math.ohio-state.edu>
Subject: #22 Contents: J. Approx. Th. - nov99 TOC

Table of Contents: J. Approx. Theory, Volume 101, Number 1, November 1999

G. G\'at 
Pointwise convergence of the Fej\'er means of functions on unbounded
   Vilenkin groups 
1--36

Daniel Su\'arez 
A generalization of the Laguerre-P\'olya class of entire functions
37--48

Yongping Liu and Guozhen Lu 
Simultaneous approximations for functions in Sobolev spaces by
   derivatives of polyharmonic cardinal splines 
49--62

T. Dunker, W. Linde, T. K\"uhn, and M. A. Lifshits 
Metric entropy of integration operators and small ball probabilities for
   the Brownian sheet 
63--77

F. Centrone and A. Martellotti 
Proximinal subspaces of $C(Q)$ of finite codimension 
78--91

Dimiter Dryanov and Qazi Ibadur Rahman 
On certain mean values of polynomials on the unit interval 
92--120

Yasuo Kageyama 
A new class of modified Bernstein operators 
121--147

G. Hal\'asz 
Markov-type inequalities for polynomials with restricted zeros 
148--155

Notes

Jes\'us de la Cal and Javier C\'arcamo 
Preservation of $p$-continuity by Bernstein-type operators 
156--159

Daniel Wulbert 
An approximation property and the sphere of $L_1$
160--163


Date: Fri, 21 Jan 2000 13:34:57 -0500 (EST)
From: Paul Nevai <nevai@math.ohio-state.edu>
Subject: #23 Contents: J. Approx. Th. - nov99 TOC

Table of Contents: J. Approx. Theory, Volume 101, Number 1, November 1999

G. G\'at 
Pointwise convergence of the Fej\'er means of functions on unbounded
   Vilenkin groups 
1--36

Daniel Su\'arez 
A generalization of the Laguerre-P\'olya class of entire functions
37--48

Yongping Liu and Guozhen Lu 
Simultaneous approximations for functions in Sobolev spaces by
   derivatives of polyharmonic cardinal splines 
49--62

T. Dunker, W. Linde, T. K\"uhn, and M. A. Lifshits 
Metric entropy of integration operators and small ball probabilities for
   the Brownian sheet 
63--77

F. Centrone and A. Martellotti 
Proximinal subspaces of $C(Q)$ of finite codimension 
78--91

Dimiter Dryanov and Qazi Ibadur Rahman 
On certain mean values of polynomials on the unit interval 
92--120

Yasuo Kageyama 
A new class of modified Bernstein operators 
121--147

G. Hal\'asz 
Markov-type inequalities for polynomials with restricted zeros 
148--155

Notes

Jes\'us de la Cal and Javier C\'arcamo 
Preservation of $p$-continuity by Bernstein-type operators 
156--159

Daniel Wulbert 
An approximation property and the sphere of $L_1$
160--163


Date: Mon, 24 Jan 2000 17:17:51 -0500 (EST)
From: Thomas Hogan <hogan@math.ohio-state.edu>
Subject: #24 Contents: J. Approx. Th. - dec99 TOC

Table of Contents: J. Approx. Theory, Volume 101, Number 2, December 1999

Leonid Golinskii 
The Christoffel function for orthogonal polynomials on a circular arc 
165--174

G. Godefroy and V. Indumathi 
Proximinality in subspaces of $c_0$
175--181

F. Schipp and W. R. Wade 
Zak transforms on binary fields 
182--195

Thomas Bloom and Jerzy Szczepaski 
On the zeros of polynomials of best approximation 
196--211

Natalia Humphreys and Ranko Bojanic 
Rate of convergence for the absolutely $(C,\alpha)$ summable Fourier
   series of functions of bounded variation 
212--220

Wolfgang Gehlen 
On a conjecture concerning strong unicity constants 
221--239

A. Kamaly, A. M. Stokolos, and W. Trebels 
On the rate of almost everywhere convergence of certain classical
   integral means, II 
240--264

Dieter Gaier 
The Faber operator and its boundedness 
265--277

Manfred Reimer 
Well conditioned spherical polynomial systems 
278--288

L. Dalla and V. Drakopoulos 
On the parameter identification problem in the plane and the polar
   fractal interpolation functions 
289--302

Alexei Zhedanov 
Biorthogonal rational functions and the generalized eigenvalue problem
303--329

T. Kasuga and R. Sakai 
Uniform or mean convergence of Hermite-Fej\'er interpolation of
   higher order for Freud weights 
330--358

Stefan Jansche 
Weak type inequalities for best simultaneous approximation in
   Banach spaces 
359--403

Author index for Volume 101 
404


Date: Mon, 31 Jan 2000 16:34:55 +0100
From: "Listowner" <listowner@baltzer.nl>
Subject: #25 Contents: Advances in Computational Mathematics 12  (2000)

Special Issue: High dimensional integration

Editor: E. Novak

Erich Novak
High dimensional integration 1-2

Leszek Plaskota
The exponent of discrepancy of sparse grids is at least 2.1933 3-24

F.J. Hickernell and H. Wozniakowski
Integration and approximation in arbitrary dimensions 25-58

Muni V. Reddy and Stephen Joe
An average discrepancy for optimal vertex-modified number-theoretic rules
59-69

Knut Petras
On the Smolyak cubature error for analytic functions 71-93

 Advances in Computational Mathematics 12  (2000) 2,3

Advances in Computational Mathematics 12 (2000) 2,3

H.N. Mhaskar and J. Prestin
On the detection of singularities of a periodic function 95-131

Serena Morigi and Marian Neamtu
Some results for a class of generalized polynomials 133-149

Kendall Atkinson and David Da-Kwun Chien
A fast matrix--vector multiplication method for solving the radiosity
equation 151-174

Michael S. Floater and Tom Lyche
Asymptotic convergence of degree-raising 175-187

Qingtang Jiang
Parameterization of m-channel orthogonal multifilter banks 189-211

V.N. Temlyakov
Weak greedy algorithms[*]This research was supported by National Science
Foundation Grant DMS 9970326 and by ONR Grant N00014-96-1-1003. 213-227

P.J. van der Houwen and B.P. Sommeijer
Diagonally implicit Runge--Kutta methods for 3D shallow water applications
229-250

Bryan P. Rynne
Convergence of Galerkin method solutions of the integral equation for thin
wire antennas 251-259

Yongdeok Kim and Sungyun Lee
Modified Mini finite element for the Stokes problem in R2 or R3 261-272



Date: Tue, 1 Feb 2000 11:51:00 -0500 (EST)
From: Thomas Hogan <hogan@math.ohio-state.edu>
Subject: #26 Contents: J. Approx. Th. - jan00 TOC

Table of Contents: J. Approx. Theory, Volume 102, Number 1, January 2000

Xiao-Ming Zeng and Wenzhong Chen 
On the rate of convergence of the generalized Durrmeyer type operators
   for functions of bounded variation 
1--12

Peter Renaud 
A Korovkin theorem for abstract Lebesgue spaces 
13--20

Ferenc Weisz 
The maximal Riesz operator of two-dimensional Fourier transforms and
   Fourier series on $H_p({\bf R}\times{\bf R})$ and
   $H_p({\bf T}\times{\bf T})$
21--45

Ding-Xuan Zhou 
Multiple refinable Hermite interpolants 
46--71

A. Kro\'o and J. Szabados 
Markov-Bernstein type inequalities for multivariate polynomials on
   sets with cusps 
72--95

Franz Peherstorfer and Robert Steinbauer 
Orthogonal polynomials on the circumference and arcs of the
   circumference 
96--119

P. D. Dragnev and E. B. Saff 
A problem in potential theory and zero asymptotics of Krawtchouk
   polynomials 
120--140

A. K. Kushpel and J. Levesley 
Quasi-interpolation on the 2-sphere using radial polynomials 
141--154

Jiang Yanjie and Liu Yongping 
Average widths and optimal recovery of multivariate Besov classes in
   $L_p(R^d)$
155--170

Note

Zhongkai Li 
Bernstein polynomials and modulus of continuity 
171--174


Date: Thu, 30 Dec 1999 14:20:22 +0800
From: Terence Sim <tsim@cs.cmu.edu>
Subject: #27 Answer: Image decompression artifacts (WD 8.10 #19)

Steve,

Try doing your compression in the RGB color space instead of YUV.
(What reason did you have for using YUV in the first place?)

In YUV space, your thresholding step effectively alters the perceived
= color, since you can't set each component to 0 independently.
Reason: = in YUV space, you lose the "orthogonality" of the color
components, so = you cannot independently threshold each component.
This is not a = problem in RGB space: if a red component is below the
threshold (i.e. = close to 0), then setting it to 0 doesn't alter the
perceived color as = much.

If you MUST work in YUV space, then you need another scheme to
quantize = the colors, taking into account that you can't
independently change each = component.  You can probably devise such a
scheme yourself.

Terence Sim



Date: Mon, 3 Jan 2000 16:05:15 +0800
From: "lin jing" <lj@farad.ioa.ac.cn>
Subject: #28 Answer: Signal classification by wavelets (WD 8.9 #27)

Dear Nan-Sheng Huang,

I think the method proposed by Yeffry Handoko may be available. The
spirit of the method is some like the "thresholding denoising"
proposed by Donoho and his colaborers. But there still have further
work to do The most important thing is the selection of the wavelet
bases. It depends on the characters of the signal. Different wavelet
bases will correspond to different wavelet coefficients. Thus, for
different wavelet bases, the biggest coefficients have different
senses. If the wavelet base is improper, the biggest coefficient will
not carry the feature information of the signal.

And I think another way is also available. You can combine the method of 
blind seperation with wavelet transform. But this method demands the
statistic property be known.

Regards,

Lin Jing
lj@farad.ioa.ac.cn
State Ley Laboratory of Acoustics
Institute of Acoustics
Academia Sinica,
Beijing, 100080
P.R.China



Date: Mon, 10 Jan 2000 11:31:12 +0000
From: "Steven Fenton" <steve.fenton@datong.co.uk>
Subject: #29 Answer: Image decompression artifacts (WD 8.10 #19)

Dear All,

Further to posting a question regarding colour artifacts that were
generated during the decompression stage of my algorithms, I received
many email's with suggesting possible solutions to the problem.

I had solved the problem shortly before receiving any email's but must
thank those who sent me their suggestions. This solution was sent to
me by Marcel Dasen of ETH Swiss Federal Institute of Technology email:
<dasen@tik.ee.ethz.ch> which correctly describes the solution.

The problem encountered was due to the YUV->RGB back-conversion. If
you have a look at the conversion matrix you will see that it is
indeed possible that the values will be off-range, when a quatized
input is used.  Since this must be considered a "rounding" error two
choices are available :
- use higher precision RGB sample values (Not possible in this application)
- use saturation (meaning values > 255 in each color channel should be
   set to 255 and values < 0 to 0.)

Regards,

Steve Fenton

Steven.Fenton@Datong.co.uk 

Steven Fenton.
Senior Design Engineer.
Datong Electronics Limited,
Clayton Wood Close, West Park,
Leeds, LS16 6QE.

TEL : 01132 744 822
FAX : 01132 742 872



Date: Mon, 17 Jan 2000 16:11:23 +0000
From: "Steven Fenton" <steve.fenton@datong.co.uk>
Subject: #30 Answer: Integer based wavelets and lifting (WD 8.10 #23)

Dear All,

Further to posting this question and in response to answers that I gratefully received. I have come to the following conclusions :

1. If I used an integer lifting scheme, which set of coefficients
   would gives the best results, with respect to the Daub4 filter I am
   currently using ?

A. S. Lewis, G. Knowles, "VLSI Architecture for 2-D Daubechies Wavelet Transform without Multipliers", Electronic Letters, Vol. 27, No. 2, 17 January 1991. In it, the authors
talk about a 2-D DWT using approximate 4-tap Daubechies coefficients of h ( 11, 19, 5, -3). Dividing each of these by 32 (a left shift of 5 positions) gives approximate floating
point coefficients.

2.  Do integer lifting schemes / Integer transforms create any
    'artifacts' not found using floating point arithmetic ?

Additional artifacts can be a result of quantization noise induced
during the transform process. In order to reduce the noise the input
values can be scaled up prior to the wavelet transforms taking
place. This, however, increases the dynamic range of the final
coefficient values. This has to be taken into account when using a
fixed point processor, particularly when baring in mind that the
coefficient values increase in magnitude each time an additional level
of transform is performed.  There are method's that do not cause the
increase in dynamic range : Chao-Fischer, An approach to fast integer
reversible wavelet transform of image compression.

3. Where can I find information specific to implementing a wavelet
   transform on a fixed point processor ?

I have not yet come across any examples. However, the integer based approaches are implicit.

Regards.

Steve Fenton

Steven.Fenton@Datong.co.uk 



Date: Fri, 31 Dec 1999 13:45:39 +0700 (JAVT)
From: Iwan Arifianto <iwan96@eng.ui.ac.id>
Subject: #31 Question: How make the model of currency forecast

Dear wavelet friends,
   First i want to give  appreciating about this mailing list. Recently, i
am doing my thesis about the expectation of the currency value with use
wavelet transform. Based on my reference, i have to make some assumption
that relate on the condition in my country (indonesia). Certainly, i have
to make a mathematics equation that represent " graphics "/signal related
with the currency value in time range.From here, I still don't understand
how make the coefficient from mother wavelet which related with my
assumption. I wish my letter can be understood by you, and please give and
sent your letter to me.

From: Iwan Arifianto  
      Electrical Engineering in University of Indonesia
      iwan96@yexa.eng.ui.ac.id 



Date: Fri, 31 Dec 1999 13:37:41 +0530
From: "K N Pratima" <chen@ad1.vsnl.net.in>
Subject: #32 Question: Applying WT to Binary Image Data

    I am currently working on using Wavelet Transform for image data =
compression. The input data is an image in binary form.I want to know =
how to make the input compatible with the Matlab code I intend to write. =
I plan to use the EZW algorithm for compressing the data.=20

Any suggestions and pointers to published material will be greatly =
appreciated.

Thank you,

K.N.Pratima
email: pratima@angelfire.com
    <knpratima@ieee.org>



Date: Mon, 03 Jan 2000 09:31:07 -0800
From: "Kevin W. Bowman" <Kevin.Bowman@jpl.nasa.gov>
Subject: #33 Question: Looking for wavelet software in C or C++

I am looking for wavelet software in C or C++ that can perform a
biorthogonal wavelet decomposition on non-dyadic vectors with
symmetric extension boundary conditions.  Ideally this software is
optimized for speed and are comparable to the computational times of a
typical FFT.  JPL would be willing to purchase this software if it
satisfies the above requirements.  For more information please contact
me at

kevin.bowman@jpl.nasa.gov

Regards,

Kevin W. Bowman, PhD.
Earth and Space Sciences Division
Jet Propulsion Laboratory
4800 Oak Grove Drive, MS 183-301
Pasadena, CA 91109



Date: Thu, 6 Jan 2000 13:14:19 -0600 (CST)
From: liyu zhu <lzh4783@cis.ksu.edu>
Subject: #34 Question: Wavelet ask for help 

My objective is to detect defects in PCB (printed circuit board),

My step is that: 

1) compare image of PCB detected with the standard PCB image. I got one  
new wimage which only has the the differnece between false PCB and good
PCB.

2) I want to use the wavelet to record and store features of many
kinds of defects.

the problem is that each kind of defect can be belong to the same kind of
defect but their size or (scale) may be differnt. and contour may be      
also little differnce althought those defect is the same kind.

How to represent those defect by wavelet?

Please give me some suggection and ideas! 

Thank you for your help! Looking forward to your reply soon!

Regards,
Zhu LiYu



Date: Sun, 16 Jan 2000 22:46:03 +0800
From: Yang Chi <yangchi@ihw.com.cn>
Subject: #35 Question: Looking for earlier versions of WavBox.

I'm a Chinese PhD student of meteorology. Signal processing is not my
specialty, but now I depend heavily on time-frequency analysis
methods, especially the MATCHING PURSUIT, in preparing for my
dissertation. I've learned from the wavelet digest_02.18 that tha
WavBox can really meet my need. But its original FTP directory is now
empty, and the newest version of wavbox is no longer a freeware. It's
not realistic for me to pay for it by credit card or check from
China. So, if anyone has earlier versions of WavBox (1.0-3.0), would
you please share it with me? I really appreciate any kind help.



Date: Fri, 04 Feb 2000 06:39:41 GMT
From: "ron kluwelder" <kluwelder@hotmail.com>
Subject: #36 Question: request for information

Let me introduce myself as a novice experimenter who wants to use
the  continuous  wavelet transform to  study  flow  visualisation
images  and  single-point sampled time series data  from  complex
flows.

For  this purpose I need to know the following. I would  be  very
grateful  to you if you could spare a minute and give me  answers
to these questions.

1.  Is there a good algorithm for 1-D and 2-D continuous  wavelet
transforms available for me to use?

2.  Whom may I contact to get the above and their address and  e-
mail, if available?

3.  Can I request for a copy of the program or do I have  to  buy
it? If I have to buy it, how much does it cost?

4. What language is this written in?

5. On what computer platform does this program work?

6.  How  much time does this algorithm take to  compute  the  1-D
continuous  wavelet transform for 5000 data points for about  500
continuous scales starting from the smallest possible scale  (say
a wavelet of 20 data points)? On what computer has this time been
achieved.

7.  How  much  time does this algorithm take  to  compute  a  2-D
continuous  wavelet transform for a 1000 x 1000  digitised  pixel
image using a wavelet of size 500 x 500 pixels? On what  computer
has this time been achieved.

8. What are the wavelets available for these computations?

With many thanks,

Ron H. Kuwelder
Visiting Fellow
Univ. of California

<webmaster@wavelet.org>