Morphology Digest July 13, 1993 1993 - Issue 1
___ _/ ____/ ____/ ____/ __ / ____/ _/ ____/ ___/ __ _
|| \ / | // \ || | || | || | // \ || // \ // \\ /
|| | | || | || | || | ||___| || | || || | || __ \\ /
|| | || | ||_/ ||_/ || | || | || || | || \ ||
|| | \\___/ || \ || || | \\___/ ||____ \\___/ \\___/ ||
Subscribe: email to morpho@cwi.nl with "subscribe" as subject
and empty message body.
Unsubscribe: email to morpho@cwi.nl with "unsubscribe" + email address
as subject and empty message body.
Submissions: email to morpho@cwi.nl with "submit" as subject.
Archive site: anonymous FTP on ftp.cwi.nl; directory /pub/morphology/digest
Editor: Henk Heijmans
-------------------------------------------------------------------------------
CONTENTS:
1. Editor's note
2. Call for papers: STOCHASTIC GEOMETRY AND STATISTICAL APPLICATIONS
3. Call for papers: 12th ICPR - SIGNAL PROCESSING
4. Paper
5. PhD thesis available
-------------------------------------------------------------------------------
1. Editor's note
-------------------------------------------------------------------------------
This is the first issue of the "Morphology Digest" to be distributed
approximately once per month. We would like to thank you for your
replies to the initial announcement and for your encouraging remarks.
This digest is intended as a forum between workers in mathematical
morphology and related fields (stochastic geometry, random set theory,
image algebra, nonlinear filtering, etc.).It can only survive with the
aid of all subscribers. So if you have any information which you consider
of interest to (part of) our readership, please submit it to this digest.
At the moment of writing there are about 300 subscribers so your submission
reaches quite a few people.
Your contribution may concern (but is not restricted to):
- conferences, workshops, courses
- books, articles, reports, PhD theses
- algorithms, software, hardware
- available research positions
- bibliographical data (e.g. bibtex files)
It is not our intention to include full reports or tex-files, but merely
to communicate title (and abstract), author(s), report number, along with
information how to obtain a copy (either by direct mail or by FTP).
We hope you all enjoy the Morphology Digest.
Henk Heijmans
Centre for Mathematics and Computer Science (CWI)
Kruislaan 413
NL 1098 SJ Amsterdam, The Netherlands
-------------------------------------------------------------------------------
2. Call for papers: STOCHASTIC GEOMETRY AND STATISTICAL APPLICATIONS
-------------------------------------------------------------------------------
From: adrianb@cwi.nl (Adrian Baddeley)
CALL FOR PAPERS
The Applied Probability Trust announces the launch of a new section
of the journal ``Advances in Applied Probability'':
STOCHASTIC GEOMETRY AND STATISTICAL APPLICATIONS.
Aims and scope.
The exciting and rapidly developing area known as ``Stochastic
Geometry'' covers a whole variety of interconnected topics, whose
common thread is the study of stochastic objects with a strongly
geometric flavour, extending from pure-mathematical studies in
integral geometry and random set theory, through topics of classical
probability such as geometric coverage problems or random processes of
geometric objects, to spatial statistics and investigations
into image analysis with a strong statistical component. These
topics include:
integral geometry,
random sets, random measures,
planar and spatial point processes, including the theory of
processes of geometric objects,
mathematical morphology,
statistical image analysis,
spatial statistics,
geostatistics,
random convex hulls,
applications of random fractal theory,
theoretical and applied statistical shape theory,
spatial bootstrapping,
edge-effects for spatial processes,
spatial limit theorems,
coverage processes,
random search algorithms,
probabilistic algorithms for computational geometry,
spatial medians.
Moreover there are now arising strong interactions between ideas of
stochastic geometry and statistics, for example in relations between the
mathematical morphology approach to image analysis and stochastic geometry,
between censoring and edge-effects, between curve fitting and Palm
measures, between bootstrapping and spatial data, between coverage models and
queuing theory.
>From March 1994 the quarterly periodical ``Advances in Applied
Probability'' (AAP) will include in each issue a new section entitled
``Stochastic Geometry and Statistical Applications'' (SGSA), to serve
as the forum for publication of research papers in these and other
topics in statistics and probability which exhibit the interplay
between geometry and probability and statistics which characterizes the
area. Papers published in SGSA will automatically reach the wide
audience of subscribers to AAP, which includes many of the world
community of stochastic geometers. AAP is also a source of reference
for many researchers seeking techniques for their applied probability
problems. Consequently the new section will not only serve as a natural
home for articles on stochastic geometry, but also is likely to promote
serendipitous and unexpected interactions with other areas of applied
probability. It is the intention of the editors that SGSA will provide
a home for the whole family of stochastic geometry topics, both
theoretical and applied, both established and innovative.
Papers should be submitted to the Sheffield office of the Applied
Probability Trust(*), following the same procedure as for AAP, but may be
marked by prospective authors as submitted specifically for publication in
SGSA. Papers for SGSA will be handled by a separate group of Editors
(including AJ Baddeley, WS Kendall, D Stoyan, RA Vitale) coordinated by WS
Kendall, the Coordinating Editor responsible for SGSA.
(*) Executive Editor
Probability and Statistics
Department of Probability and Statistics
The University
Sheffield S3 7RH, UK
-------------------------------------------------------------------------------
3. Call for papers: 12th ICPR - SIGNAL PROCESSING
-------------------------------------------------------------------------------
From:malah@ee.technion.ac.il (David Malah)
CALL FOR PAPERS - 12th ICPR - SIGNAL PROCESSING
Oct 9-13, 1994, Jerusalem, Israel
CONFERENCE TOPICS:
Analysis, representation, coding, recognition and enhancement of signals:
speech and image enhancement; image restoration; speech and audio coding;
image and video coding; multiresolution and wavelet analysis; scale-space
and joint time-frequency representation; morphological signal processing;
auditory scene analysis.
PROGRAM COMMITTEE:
D. Malah (Chair) - Technion (malah@techunix.technion.ac.il)
M. Barlaud J. Biemond A. Bregman
Y. Bresler D. Chazan A. Cohen
V. Cuperman Y. Ephraim S. Furui
A. Gill N. Jayant E. Karnin
M. Kunt S. Mallat W. Pearlman
A. Rosenberg Y. Sakai P. Salembier
M. Tekalp Y. Yasuda Y. Zeevi
This conference is one of Four conferences in the 12th ICPR. Each submitted
paper will be reviewed by members of the program committee.
Papers describing applications are encouraged, and will be reviewed
by a special Applications Committee.
The conference proceedings are published by the IEEE Computer Society Press.
12-ICPR CO-CHAIRS: S. Ullman - Weizmann Inst. (shimon@wisdom.weizmann.ac.il)
S. Peleg - The Hebrew University (peleg@cs.huji.ac.il)
LOCAL ARRANGEMENTS: Y. Yeshurun - Tel-Aviv University (hezy@math.tau.ac.il)
INDUSTRIAL & APPLICATIONS LIAISON: M. Ejiri - Hitachi (ejiri@crl.hitachi.co.jp)
PAPER SUBMISSION DEADLINE: February 1, 1994.
Notification of Acceptance: May 15, 1994. Camera-Ready Copy: June 30, 1994.
Send four copies of paper to: 12th ICPR, c/o International, 10 Rothschild Blvd,
65121 Tel Aviv, ISRAEL. Tel. +972(3)510-2538, Fax +972(3)660-604
Each manuscript should include the following:
1. A Summary Page addressing these topics:
- Indicate that the paper is submitted to the SIGNAL PROCESSING conference.
- What is the paper about? - What is the original contribution of this work?
- Does the paper mainly describe an application, and should be reviewed by
the applications committee?
2. The paper, limited in length to 4000 words. This is the estimated length
of the proceedings version.
For further information on all ICPR conferences contact the secretariat at the
above address, or use E-mail: icpr@math.tau.ac.il .
-------------------------------------------------------------------------------
4. Paper
-------------------------------------------------------------------------------
From: adr@ccr.jussieu.fr (Adrian Raftery)
Banfield, J.D. and Raftery, A.E. (1992). "Automatic identification
of ice floes using mathematical morphology and clustering about
principal curves." Journal of the American Statistical Association
87, 7-16.
-------------------------------------------------------------------------------
5. PhD thesis available
-------------------------------------------------------------------------------
From: rein@fwi.uva.nl (Rein van den Boomgaard)
Last year I received my PhD degree. The thesis is entitled
"Mathematical Morphology: Extensions towards Computer Vision". Some
copies of the thesis are still available. If you send me an e-mail I
will be glad to send a copy by surface mail. The table of contents and
the summary are given below.
Rein van den Boomgaard
Faculty of Mathematics and Computer Science
University of Amsterdam
Kruislaan 403
1098 SJ Amsterdam
The Netherlands
email: rein@fwi.uva.nl
=================
TABLE OF CONTENTS
=================
1 An Introduction to Mathematical Morphology
PART 1: DECOMPOSITION OF STRUCTURING ELEMENTS
2 Logarithmic Decomposition of Convex Sets
3 Dimensional Decompositions of Functions
4 Methods for Fast Morphological Image Processing
PART II: THRESHOLD MORPHOLOGY
5 Threshold Logic and Mathematical Morphology
6 Threshold Morphology for Functions
7 Self Learning Image Processing
PART III: TOWARDS COMPUTER VISION
8 Morphological Propagators: The Differential Equations
of Mathematical Morphology
This thesis is divided in three parts. Part I entitled ``Decomposition
of Structuring Elements'' extends the classical theory of mathematical
morphology as it deals with the computational aspects of implementing
the basic morphological transformations. In chapter 2 we look at an
efficient method to decompose a large convex structurng elements into
a small number of sets. The logarithmic decomposition of sets
provides us with a tool to efficiently implement erosions and
dilations using large structuring elements.
It is a well known result from the linear theory that the unique
2-dimensional isotropic convolution which can be dimensionally
decomposed in a horizontal convolution followed by a vertical
convolution, is the convolution using a Gaussian convolution mask.
This property is an important one from a practical point of view as it
reduces the computational complexity of a convolution from NxN to 2xN
where NxN is the size of the convolution mask. Chapter 3 introduces
the ``morphological equivalent'' of the Gaussian filter. There we
prove that the Quadratic Structuring Functions have analog properties
under erosions and dilations.
Chapter 4 concludes part I with a the description of algorithms to
implement the basic morphological operations efficiently on general
purpose computers.
The basic operations defined in classical mathematical morphology (the
erosion and dilation) are sensitive to noise because erosion and
dilation are defined in terms of set equality. Only one noise pixel
thus changes the outcome of both the erosion and dilation. The use of
relatively large \se s (with a high probability to ``hit'' a noise
pixel) to detect geometrical structures in images renders the use of
erosions and dilations virtually useless. In part II we introduce
threshold morphology to effectively deal with noise and/or uncertainty
in the image and/or model. Chapter 5 describes threshold morphology
for sets. A subset of these threshold transforms are then generalized
to work with functions (grey value images) in chapter 6. Threshold
morphology provides the formalism to analyze, for instance, the
well-known percentile filters from within a morphological framework.
Threshold transforms bear great resemblance with the perceptrons and
adaptive linear elements, which are nowadays considered to be relative
simple examples of neural networks. In chapter 7 a learning procedure
is described based on the fact that neighbouring pixels in an image
are strongly correlated. The use of this a priori knowledge about
spatial coherence in images results in much faster converging learning
procedures.
The final part of this thesis is entitled ``Towards Computer Vision'',
and deals with the analysis of dilations using a structuring element
of variable size. As the size (scale) of the structuring element is
increased the dilation result changes. We show that we can link an
infinitesimal small change in scale with an infinitesimal change in
the dilation result. The resulting differential equations provide the
theoretical tools to investigate morphological scale-space.
This final chapter again shows that mathematical has wider
applicability than the fields of image processing and computer vision.
The differential equations solved with morphological propagators are
well-known in physics as the conservation laws describing the
propagation of shock-waves. In image processing, these differential
equations govern the construction of morphological scale-spaces.
-------------------------------------------------------------------------------
End of Morphology Digest
-------------------------------------------------------------------------------