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

The Wavelet Digest - Volume 2, Issue 9 (June 18, 1993)


Today's Editor: Wim Sweldens
                sweldens@math.sc.edu


Today's Topics:
    1. Thesis available: Image Processing with Multiscale Stochastic Models
    2. Preprint available: Wavelet Electrodynamics, Part II
    3. New Journal: Electronic Journal of Differential Equations (EJDE)
    4. International conference on wavelets, Taormina: preliminary program
    5. Question: Analysis of Nystagmus Data Using Wavelets
    6. Wavelet literature survey available
    7. Thesis available: Studies in Turbulence using Wavelet Transforms for Data Compression and Scale-Separation
    8. Cfp: 1994 Intern. Symp. on Speech, Image Processing & Neural Networks
    9. Thesis available: Applications of Wavelets to Quantization and Random Process Representations
   10. Question: boundary problems with wavelets

Submissions for Wavelet Digest:
    E-mail to wavelet@math.sc.edu with "submit" as subject.

Subscriptions for Wavelet Digest:
    E-mail to wavelet@math.sc.edu with "subscribe" as subject.
    To unsubscribe, e-mail with "unsubscribe" followed by your e-mail
    address as subject. To change address, unsubscribe and resubscribe.

Archive site and back issues:
    Anonymous ftp to maxwell.math.sc.edu (129.252.12.3),
    directory /pub/wavelet/archive.

Gopher server: bigcheese.math.sc.edu.

 
Current number of subscribers: 2704


From: Mark Luettgen, MIT Subject: PhD thesis abstract Author: Mark R. Luettgen Advisor: Prof. Alan S. Willsky Title: Image Processing with Multiscale Stochastic Models Institution: Massachusetts Institute of Technology Copies: This thesis can be obtained by writing to: MIT Laboratory for Information and Decision Systems Publications Office Room 35-311 77 Massachusetts Ave. Cambridge, MA 02139 or via anonymous ftp from: machine name: lids.mit.edu, directory: pub/ssg/papers Abstract: In this thesis, we develop image processing algorithms and applications for a particular class of multiscale stochastic models. First, we provide background on the model class, including a discussion of its relationship to wavelet transforms and the details of a two-sweep algorithm for estimation. A multiscale model for the error process associated with this algorithm is derived. Next, we illustrate how the multiscale models can be used in the context of regularizing ill-posed inverse problems and demonstrate the substantial computational savings that such an approach offers. Several novel features of the approach are developed including a technique for choosing the optimal resolution at which to recover the object of interest. Next, we show that this class of models contains other widely used classes of statistical models including 1-D Markov processes and 2-D Markov random fields, and we propose a class of multiscale models for approximately representing Gaussian Markov random fields. These results, coupled with those illustrating the computational efficiencies that the multiscale models lead to, suggest that the multiscale framework is a powerful paradigm for image processing both because of the efficient algorithms it admits and because of the rich class of phenomena it can be used to describe. This motivates us in the final section of this thesis to pursue further algorithmic development for the multiscale models. In particular, we develop an efficient likelihood calculation algorithm for multiscale models and demonstrate an application of the algorithm in the area of texture discrimination. The thesis concludes with a review of our main results and with a discussion of a few of the many open problems and promising directions for further research and application.
From: Gerald Kaiser, University of Massachusetts at Lowell. Subject: Preprint available: Wavelet Electrodynamics, Part II *** this is a tex-file, cut here *** \magnification=1200 \baselineskip=14pt \hsize=16 true cm \vsize=21 true cm \def\v#1{{\bf#1}} \def\c#1{{\cal#1}}\def\t#1{\tilde#1} \centerline{\bf WAVELET ELECTRODYNAMICS, Part II:} \vglue 1ex \centerline{\bf Electromagnetic Wavelets with Polarization} \vglue 3ex \centerline{Gerald Kaiser} \centerline{Department of Mathematics} \centerline{University of Massachusetts at Lowell} \centerline{Lowell, MA 01854, USA} \centerline{e--mail: kaiserg@ woods.ulowell.edu} \vglue 2ex \centerline{June 2, 1993} \vglue 2ex \centerline{\bf ABSTRACT} \vglue 1ex \noindent The representation of free electromagnetic waves (solutions of Maxwell's equations) as superpositions of scalar wavelets with vector coefficients developed earlier is generalized to wavelets with polarization. The construction is canonical and proceeds in four stages: {\bf(1)} The {\sl analytic--signal transform} is used to extend fields from real spacetime to a tube domain $\c T$ in complex spacetime. {\bf(2)} The {\sl evaluation maps,\/} which send any field $\v F$ to the values $\t\v F(z)$ of its extension at points $z\in\c T$, are bounded linear maps on the space of solutions. Their adjoints $\v W_z$ are the electromagnetic wavelets. These are matrix--valued to include all allowed states of polarization. {\bf(3)} The eight real parameters $z=x+iy\in \c T$ are given a complete physical interpretation: $x=(\v x, t)\in\v R^4$ is interpreted as a spacetime point about which $\v W_z$ is {\sl focussed,\/} i.e., $\v W_{x+iy}$ is localized around the space point $\v x$ at time $t$. The imaginary spacetime variable $y=(\v y, s)$ is timelike. Its Lorentz--invariant magnitude $\lambda=(s^2-|\v y|^2)^{1/2}$ is interpreted as the {\sl scale} of the wavelet (i.e., its width at the time of maximal focus), and its direction $\v v=\v y/s$ as the {\sl velocity} of its center. In particular, wavelets $\v W_z$ parameterized by {\sl Euclidean\/} points $z=(\v x, is)$ (real space and imaginary time) have stationary centers; the real space point $\v x$ is their point of localization, and the imaginary time is their scale. The set $E\subset\c T$ of all such points is accordingly called {\sl wavelet space.} {\bf(4)} A resolution of unity is derived for the Hilbert space of solutions which gives a representation of any solution $\v F\in \c H$ as a superposition of the wavelets $\v W_z$ parameterized by $z\in E$. \bye
From: Julio G. Dix, Rice University Subject: New Journal: Electronic Journal of Differential Equations (EJDE) ELECTRONIC JOURNAL OF DIFFERENTIAL EQUATIONS (EJDE) Mathematicians at Southwest Texas State University and at the University of North Texas have collaborated to establish a new journal, the ELECTRONIC JOURNAL OF DIFFERENTIAL EQUATIONS (EJDE). The EJDE will be a strictly electronic journal dealing with all aspects of differential equations. Articles will be submitted as TeX files, sent to referees electronically, and then disseminated electronically, free of charge. Although the time between submission and dissemination will be greatly reduced, only original research of high quality will be accepted. Each article will be subject to as rigid a peer review process as is applied by the finest of today's printed journals. The EJDE is calling for papers now. There are no page charges. The EJDE can be reached via: 1. "telnet ejde.math.unt.edu", login: "gopher" 2. "telnet e-math.ams.com", login: "e-math", password: "e-math", select line 12, then line 14 3. "ftp ejde.math.unt.edu" 4. "gopher ejde.math.unt.edu", if the gopher client software is loaded in your computer. As soon as the second file server arrives, the EJDE will also be available at the above addresses replacing "unt" by "swt". Readers can transfer the TeX, DVI and Postcript files to their own computers and then read them or print hard copies. Identical copies of the journal will be maintained for posterity and for interlibrary loan at the libraries of the two universities. A free subscription to the abstracts of new articles in the EJDE is available by sending an e-mail message to "editor@ejde.math.unt.edu". The Managing Editors of EJDE are Alfonso Castro, Julio Dix, Gregory Passty, and Ricardo Torrejon. The Editorial Board consists of P. Bates (Brigham Young University) A. Bloch (Ohio State University) J. Bona (Pennsylvania State University) K. J. Brown (Heriot-Watt University) L. Caffarelli (Institute for Advanced Study) C. Castillo-Chavez (Cornell) C. Chui (Texas A & M University) M. Crandall (University of California at Santa Barbara) E. Di Benedetto (Northwestern University) G. B. Ermentrout (University of Pittsburgh) J. Escobar (Indiana University) L. C. Evans (University of California at Berkeley) J. Goldstein (Louisiana State University) C. Groetsch (University of Cincinnati) I. Herbst (University of Virginia) C. Kenig (University of Chicago) R. Kohn (Courant Institute) A. Lazer (Miami University) J. Neuberger (University of North Texas) P. H. Rabinowitz (University of Wisconsin) R. Shivaji (Mississippi State University) R. Showalter (University of Texas) H. Smith (Arizona State University) P. Souganidis (University of Wisconsin) N. Walkington (Carnegie-Mellon University)
From: L. Puccio and L. Montefusco, University of Bologna, Italy; Subject: International conference on wavelets, Taormina: preliminary program INTERNATIONAL CONFERENCE ON WAVELETS Theory, Algorithms, and Applications Taormina (Italy) - October 14-20, 1993 In collaboration with "Academia Peloritana dei Pericolanti", "Consiglio Nazionale delle Ricerche (C.N.R.)", "Gruppo Nazionale per l' Informatica Matematica del C.N.R. (G.N.I.M.)", the University of Messina and its Department of Mathematics will organize an International Conference on Wavelets at Taormina (Sicily - Italy) from October 14 to 20, 1993. Organizers L. Puccio - Department of Mathematics, University of Messina, Italy; L. Montefusco - Department of Mathematics, University of Bologna, Italy; A. Mammano - Department of Mathematics, University of Messina, Italy; G. Crupi - Department of Mathematics, University of Messina, Italy; C. K. Chui - Center for Approximation Theory, Texas A&M University, USA. Scientific Committee C. K. Chui - Texas A&M University, USA; R. Coifman - Yale University, USA; I. Daubechies - Rutgers University, USA; I. Galligani - University of Bologna, Italy; A. Mammano - University of Messina, Italy; Y. Meyer - University of Paris Dauphine, France; L. Montefusco - University of Bologna, Italy; L. Puccio - University of Messina, Italy. PROGRAM Oct. 14, Thursday 9:00 - 10:00 A. Grossmann 10:25 - 12:30 20 minute contributed papers (2 parallel sessions) 15:00 - 16:00 D. Trigiante 16:15 - 16:45 K. Jetter 16:45 - 18:50 20 minute contributed papers (2 parallel sessions) Oct. 15, Friday 9:00 - 10:00 W. Dahmen 10:25 - 12:30 20 minute contributed papers (2 parallel sessions) 15:00 - 16:00 M. V. Wickerhauser 16:15 - 16:45 R. Lorentz 16:45 - 18:50 20 minute contributed papers (2 parallel sessions) Oct. 16, Saturday 9:00 - 10:00 C. Micchelli 10:25 - 11:25 A. Cohen 11:30 - 12:30 C.K. Chui Oct. 18, Monday 9:00 - 10:00 S. Mallat 10:25 - 12:30 20 minute contributed papers (2 parallel sessions) 15:00 - 16:00 S. Jaffard 16:15 - 16:45 J. Stoeckler 16:45 - 18:50 20 minute contributed papers (2 parallel sessions) Oct. 19, Tuesday 9:00 - 10:00 M. Farge 10:25 - 12:30 20 minute contributed papers (2 parallel sessions) 15:00 - 15:30 R. Wells 15:30 - 16:00 L. Schumaker 16:05 - 18:50 20 minute contributed papers (2 parallel sessions) Oct. 20, Wednesday 9:00 - 10:00 L. Montefusco 10:30 - 11:00 E. Saff 11:00 - 12:30 20 minute contributed papers (2 parallel sessions) 15:00 - 17:30 20 minute contributed papers (2 parallel sessions) INFORMATION The Conference will be held at Jolly Hotel Diodoro located in the center of Taormina ( near "Villa Comunale" ). Deadline for abstracts: June 30, 1993 Participants who wish to submit a research talk, poster or software demonstration should send title and abstract before June 30, 1993. The abstract, typed double-spaced, of one page maximum, will be directly reproduced in the booklet handed out to all participants. Proceedings Invited papers as well as some selected contributed papers will be published by Academic Press in 1994. All papers will be refereed and carefully edited. Registration Registration fee is Lit. 320.000 until June 30, 1993. Late registration is Lit. 370.000. The registration fee will cover six lunches, all coffee breaks, the abstract book and the edited volume of conference papers. No fee is requested for accompaining persons, but tickets for lunch ( Lit. 30.000 each ) Laura Montefusco - Dipartimento di Matematica , Universita' di Bologna Piazza di Porta S. Donato, 5 - 40127 Bologna, Italy Phone: 39-(0)51-354439, Fax: 39-(0)51-354490, E-mail: MONTELAU+DM.UNIBO.IT Luigia Puccio - Dipartimento di Matematica , Universita' di Messina Salita Sperone, 31 - Contrada Papardo - 98166 Messina, Italy Phone: 39-(0)90-6763066 Fax: 39-(0)90-393502 E-mail: GINA+IMEUNIV.BITNET Secretariat G. Cusumano, M. Chui, V. Calderone, M. Cotronei, D. Lazzaro, M. Piraccini. Phone: 39-(0)90-6763068 Fax: 39-(0)90-393502 Address: International Conference on Wavelets Dipartimento di Matematica - Universita' di Messina Salita Sperone, 31 - Contrada Papardo 98166 Messina, Italy E-mail: GINA+IMEUNIV.BITNET Note from the editor: this message was shortened, for more information and registration forms please contact the organizers.
From: Redouan Rouzky, University of Texas at San Antonio Subject: Question: Analysis of Nystagmus Data Using Wavelets I would appreciate if anybody can provide me with some information (reference papers, books, ...) related to the analysis of nystagmus data using wavelets (nystagmus is a reflexive eye movement). My goal is to analyze "slow phase velocity" and "fast phase velocity" from the nystagmus data. Thanks rousky@ennex3.eng.utsa.edu
From: Stefan Pittner, Technical University Vienna Subject: Wavelet literature survey available Stefan Pittner, Josef Schneid and Christoph W. Ueberhuber WAVELET LITERATURE SURVEY Institute for Applied and Numerical Mathematics Technical University Vienna 1993, approx. 220 pages, Softcover, US$ 20 For the past two years the authors collected any major publication available on the theory of classical wavelets - translations and dilations of one or more functions. This has grown into a bibliography of more than 700 entries, 60 percent of them supplemented by abstracts. There are four categories -- books, articles, topics and authors. This up-to-date list represents valuable information on basic and state-of-the-art wavelet literature for the beginner and the scientific researcher as well. If you are interested in a copy of this reference list (which we will not pass in the form of files !), then please send an e-mail to pittner@uranus.tuwien.ac.at for more informations. You are kindly requested to pass on this e-mail to anyone potentially interested in wavelets. Best wishes Stefan Pittner
From: Lareef Zubair, Yale University Subject: Thesis available The thesis described below is available from University Microfilms as well as postscript files for ftp from the author requests to - zubair%chaos@venus.ycc.yale.edu until I find an archive where I can dump them Studies in Turbulence using Wavelet Transforms for Data Compression and Scale-Separation Lareef Zubair Abstract Fluid turbulence is characterized by localized structure of multiple scales. Wavelet transforms are novel techniques which can be used to analyze localized data with multiple scales efficiently. Motivated by this congruence, we use wavelet transform to study the structure of turbulent flows. Wavelet transform is a generic term and we use, in particular, the continuous wavelet transform and the wavelet-packet transform. We study the structure of scalar and vortical fields using the continuous wavelet transform. We assess the wavelet-packet as a tool for data compression and introduce a wavelet-packet filtering technique. We use wavelet-packet filtering to pick out features of particular bands of scales, and based on this delineation, study three facets of the structure of turbulence -- local isotropy, intermittency and the scaling of structure functions. This thesis presents an assessment of the utility of wavelet transforms as well as a new understanding of some aspects of the structure of turbulence.
From: Dr. Oscar Au, University of Hong Kong Subject: Cfp: 1994 Intern. Symp. on Speech, Image Processing & Neural Networks CALL FOR PAPER 1994 International Symposium on Speech, Image Processing & Neural Networks Organizer: IEEE Hong Kong Section, and IEEE Signal Processing Hong Kong Chapter Co-sponsor: IEEE Signal Processing Society Date: Apr. 14-16, 1994 Location: Hong Kong Convention & Exhibition Center, Hong Kong TOPICS INCLUDE: 1. Speech production, preception, coding, recognition, understanding & language modeling, enhancement and noise reduction, analysis, synthesis and speaker recognition 2. Computer vision, scene analysis, image understanding, image and video coding, modeling, wavelets, color images, feature extraction, segmentation, pattern recognition, sensor fusion, fast algorithm, and software & hardware techniques. 3. Artificial neural network theory and applications, network architecture, algorithm, adaptive resonance theory, supervised & unsupervised learning, neural fuzzy system and neuro-recognition This symposium is specially designed to be immediately before ICASSP '94 which will be held in Adelaide, Australia. Perspective authors are invited to submit four copies of extended summaries of 1000-1500 words, with illustrations if necessary, and an abstract of less than 100 words, to Dr. Chorkin Chan (Program Chairman) Department of Computer Science University of Hong Kong Pokfulam Road, Hong Kong Tel: (852) 859-7075 Fax: (852) 559-8447 Email: cchan@csd.hku.hk AUTHORS' SCHEDULE Deadline for extended summaries: Oct. 4, 1993. Notification of acceptance: Dec. 20, 1993. Submission of camera-ready papers: Jan. 31, 1994. For further information, please contact Dr. Daniel Lun at enpklun@hkpcc.hkp.hk.
From: Marc Goldburg, Stanford University Subject: Thesis available THESIS ABSTRACT Title: Applications of Wavelets to Quantization and Random Process Representations Author: Marc Goldburg Date: May, 1993 Advisor: Thomas Kailath Institution: Stanford University Availability: 1) University Microfilms, or 2) via anonymous ftp to rascals.stanford.edu: 'pub/marcg/mgThesis1side.ps.Z' and 'pub/marcg/mgThesis2side.ps.Z' are 'compress'ed PostScript files suitable for one and two-sided printing, respectively. Despite its short history, the wavelet transform has found application in a remarkable diversity of disciplines: Mathematics, Physics, Numerical Analysis, Signal Processing and others. In this thesis, we examine the utility of this transform for three different signal processing applications: the representation of continuously indexed random processes; transform vector quantization systems; and partial representations and subband coding of discretely indexed random processes. For continuously indexed random processes, we show that orthogonal wavelet series representations converge in $L^2(R)$ with probability one and in mean-square for all finite energy random processes. With compactly supported wavelets and scaling functions, these same convergence modalities hold for finite power random processes on any compact interval. Between them, the finite power and finite energy random processes include all physical random processes. Next, we present a high-rate analysis of Transform Vector Quantization (TVQ) systems subject to a rate or index entropy constraint. This analysis provides expressions for the best achievable distortion and optimal sub-vector rate or index entropy assignments for a given source, orthogonal transform, and sub-vector partitioning scheme. Bounds are provided for the performance penalty incurred by TVQ relative to unstructured vector quantization. These analytical results are especially useful in the design of TVQ systems; they permit the relative performance of competing transforms to be assessed using only the source statistics. If the class of candidate transforms has a parametric description, our analytical results lead to a tractable optimization problem whose solution is the optimal (distortion minimizing) transform within the class. To demonstrate this application, the general results are specialized to the case of a Gaussian source and a transform cost function identified. The properties of transforms minimizing this cost and a computationally efficient upper bound on the cost are then presented. Using the Discrete Time Wavelet Transform (DTWT) as our parametric family of transforms, we present design examples of optimal TVQ systems and compare their performance to systems based on the Discrete Cosine Transform. The final topic addressed is the ability of a single DTWT step to concentrate the energy in its input sequence into one of its two component output subsequences. This issue relates, for example, to maximization of the coding gain afforded by scalar subband coding systems employing the DTWT. We present bounds on the concentration capability of the transform, and discuss the design of optimal (maximally concentrating) DTWT's. For any wide-sense stationary process, we show that the optimal DTWT with length four filters will always be one of two fixed choices, one of which corresponds to Daubechies' $D_2$ scaling function.
From: Daniel K R McCormack Subject: Question: Daub4 wavelets in 2dimensions I am trying to recognise an image using a wavelet encoding of a 128 x 128 pixel image. I have been led to beleive that if you take the daub4 wavelet transform of a 2D image that there are reflection and boundary problems. 1) I have been unable to find any literature on this matter - any suggestions / refs would be greatfully preciated. 2) Will these boudary / reflection problems affect the classification of the image ? Many Thanks Daniel McCOrmack Daniel.K.R.McCormack@cm.cf.ac.uk

<webmaster@wavelet.org>