Date: 27 Aug 91 10:38:22-PST
From: Vision-List moderator Phil Kahn <Vision-List-Request@ADS.COM>
Errors-to: Vision-List-Request@ADS.COM
Reply-to: Vision-List@ADS.COM
Subject: VISION-LIST digest 10.36
To: Vision-List@ADS.COM

VISION-LIST Digest    Tue Aug 27 10:38:22 PDT 91     Volume 10 : Issue 36

 - Send submissions to Vision-List@ADS.COM
 - Vision List Digest available via COMP.AI.VISION newsgroup
 - If you don't have access to COMP.AI.VISION, request list 
   membership to Vision-List-Request@ADS.COM
 - Access Vision List Archives via anonymous ftp to ADS.COM

Today's Topics:

 Neural networks for morphological operators
 Genetic algorithms and edge tracing
 Range image archive
 Colour, Texture, Fractals
 Colour Image Enhancement by Colour Manipulation
 Bar Code software
 Biomch-L
 CFP: AI, Simulation and Planning '92
 Sr. Image Understanding Scientist/Engineer Position

----------------------------------------------------------------------

Date: Thu, 22 Aug 91 17:56:53 EDT
From: cherwig@eng.clemson.edu (christoph bruno herwig)
Subject: Neural networks for morphological operators

Dear all,

I am interested in having neural networks learn morphological operators
like DILATION, EROSION, CLOSING, OPENING, SKELETONIZING.
Initial attempt is an input-layer/output-layer feedforward network and 
backprop learning rule.

I would appreciate any pointers to people who might have done research
in this area or to papers that might have been published since I
couldn't find *any*.

Thanks in advance,
Christoph Herwig

Dept. of Electrical and Computer Engineering, Clemson University, SC

------------------------------

Date: Tue, 27 Aug 91 17:09:17 GMT
From: lapena@athena.cs.uga.edu (Trurl)
Organization: University of Georgia, Athens
Subject: genetic algorithms and edge tracing
Keywords: genetic algorithms, edge tracing

Does anyone out there know of a genetic algorithm application to edge
tracing?

suggestions will do.

-chito

 lapena@athena.cs.uga.edu          [] "And if anyone doubts these figures
 clapena@aisun1.ai.uga.edu         [] then let him work it out for himself."

------------------------------

From: flynn@yoda.eecs.wsu.edu (Patrick J. Flynn)
Date: Sun, 25 Aug 91 19:35:21 PDT
Subject: range image archive
The range image archive has been re-established here at Washington State
University.  The machine name and IP address are:

gecko.eecs.wsu.edu	(134.121.32.17)

The data files live in the /pub directory tree.
Please remember to use binary mode when transferring files, and
perform ftp transfers AFTER `normal' business hours (i.e. do it
between 1700 and 0500 Pacific time).  I reserve the right to disable
ftp activity during the day if it significantly impacts my ability
to use gecko for my work.

The archive has been expanded to include CAD models (IGES and polyhedral
approximations) of twenty objects.  See pub/cad-models/README for details.
Also, several new images have been provided by Timothy Newman at Michigan State
University.  They appear in the pub/range-images/moremisc directory.  The
rest of the archive has not changed, so if you retrieved the files while
they were at Notre Dame, you don't need to get the `synth', `isolated', or
`cluttered' imagery again.

***** previous readme
This directory contains a bunch of range images produced by 

- the MSU Pattern Recognition and Image Processing Lab's Technical
  Arts 100X scanner (aka `White scanner'), OR
- scan-conversion software that I adapted from a program
  originally written by Paul Besl many years ago.

You are free to use these images to test your algorithms.  If the
images are to appear in a published article, please acknowledge the MSU
PRIP Lab as the source of the images (you don't have to mention my
name, though).

File format: rather than deal with all the goofy standards out there
for images (and to preserve the floating-point representation), these
images are compressed ASCII text files.  Beware: they expand by about
10x when uncompressed.  I recommend that you keep them compressed to
save disk space.  Many of you will probably convert these files to
your own `local' image format anyway.

Each image file has a three-line header giving the number of rows and
columns.  This is followed by four images.  The first is the so-called
'flag' image, where a pixel value of 1 means the corresponding (x,y,z)
values at that pixel are valid.  If the flag value is zero, you should
ignore the (x,y,z) components for that pixel.

Following the flag image is the image of X-coordinates, the image of
Y-coordinates, and the image of Z-coordinates.  All are floating-point
images.  MSU's White scanner is configured so that each stripe of
range values occupies one column in the image.  The object is swept
under the stripe with an XY table to get an image.  So the X
coordinate image is a linear ramp; the X value is taken from the
absolute position of the X stage in the XY table.  The Y value depends
on the column number of the pixel, and the Z value is the measured
range (the height above a table).

You can use the 3D coordinates of each range pixel, or you can
throw away the X and Y images, and concern yourself with the Z-value
alone.  Note that the `aspect ratio' of the image doesn't
have to be 1, although I try to keep it in the neighborhood of 1.

Remember to use binary mode when you transfer the images.

Contents:

In this directory, there are 4 subdirectories:

NAME            SIZE    CONTENTS
cluttered       1302KB  10 White Scanner images of pairs of objects

------------------------------

Date: Mon, 19 Aug 91 21:01:50 GMT
From: ysa@computing-maths.cardiff.ac.uk (Yuksel Samast)
Organization: University of Wales College of Cardiff, Cardiff, WALES, UK.
Subject: Colour, Texture, Fractals

       Hi Netters !

       I'm a research (Ph.D.) student working on the combination of
       colour and texture through fractal geometry in computer vision
       and this is my first mail to you all.

       I'm trying to complete a good (survey,review,report) covering
       all aspects of colour , texture , and the interaction of colour
       with the other informations such as,texture, stereo, motion,
       contour, etc...

       If anyone has anything out there,
       please do me a favour and send me whatsoever relevant.
       Papers,comments,etc...would be much appreciated !

       I will give the net all list of references when I get them.

       Many thanks in advance.

       SAMAST

       Address: Mr. Y. SAMAST
		Univ. of WALES College of CARDIFF
		ELSYM Depart.
		P.O. Box 904 , Cardiff CF1 3YH /U.K.
		Tel: (222) 874000/ext 5958
		Fax: (222) 874192
       e-mail : ysa@computing-maths.cardiff.ac.uk
		or
		samast@uk.ac.cardiff.taff

------------------------------

Date: 25 Aug 91 11:20:19 GMT
From: stuart@techunix.technion.ac.il (Stuart Wolf)
Organization: Technion, Israel Institute of Technology, Haifa, Israel
Subject: Colour Image Enhancement by Colour Manipulation
Keywords: Colour Enhancement

I am presently working on my masters thesis in which I am working
on techniques to improve the appearance of colour images by changing
subtly the colours. In this way I hope to obtain an apparent contrast
improvement thus rendering detail in the image more visible.

Very little work has been in this area and so I am appealing to anyone 
who may know of references relating to this. I am aware of the work
by Land and McCann (Retinex Theory) and also of the colour constancy
work done by Ron Gershon, Gershon Buchsbaum and Olivier Faugeras.
I am not really intersted in colour constancy so if anyone knows of
something relevant I'd be very grateful.

You may reply to (stuart@techunix.technion.ac.il)

Thanks

Stuart

------------------------------

Date: Fri, 16 Aug 1991 16:16:17 GMT
From: burback@indetech.com (Ron Burback)
Organization: Independence Technologies, Inc. Fremont, CA
Subject: Bar Code software

Hi all,

Does anyone have available, commercial or public domain, bar code
recognition software that runs under UNIX off of a fax machine
generated image?

thank you in advance for your help.

-ron
-- 
____*_  Ron Burback                 burback@indetech.com
\  / /  Independence Technologies   {sun,sharkey,pacbell}!indetech!burback
 \/ /   42705 Lawrence Place        FAX: 415 438-2034
  \/    Fremont, CA 94538           Voice: 415 438-2087

------------------------------

Date: Tue, 20 Aug 91 23:08 MET
From: "Herman J. Woltring" <UGDIST@nici.kun.nl>
Subject: Biomch-L

Dear Vision-List readers,

At the request of the moderator, it is my pleasure to post a summary
statement on Biomch-L@nic.SURFnet.NL / Biomch-L@hearn.bitnet at
Nijmegen University, The Netherlands.

Biomch-L is concerned with Biomechanics and (biological) Movement Science; as
such, the list has rather much overlap with the Vision List, in particular as
regards its current focus on 3-D kinematics and inverse dynamics.  Recent pos-
tings and discussions -- some of which quite lengthy -- have been concerned
with:
   3-D analytical photogrammetry and camera (self)calibration
   3-D rigid-body movement from single and multiple camera views
       on landmarks where the correspondence problem has been
       solved separately or is simultaneously considered <*>
   3-D definitions of instantaneous centres and axes of rotation
   3-D `helical' angle definitions avoiding the singularities and
       otherwise strongly correlated properties of Euler/Cardan angles
   3-D definitions of net joint moments (torques/couples)
   Optimal smoothing and differentiation of low-pass signal/wide-band
       noisy data using natural, quintic and heptic splines with GCV
       (Generalized Cross Validation)

   <*> e.g., a recently announced thesis from the University of Umeaa in
       Sweden.

plus discussions on their calculus.  Summary descriptions of some FORTRAN
codes implementing the above are available from LISTSERV@HEARN.BITNET via
the command INDEX BIOMCH-L; individual files (some of which are for sub-
scribers only) can then be obtained by similar commands
   SEND FILE_NAME FILE_EXTENSION
to the same address.

Postings to the list occur to the address Biomch-L@hearn.bitnet or to
Biomch-L@nic.SURFnet.NL, and subscription occurs via the one-line request

   subscribe Biomch-L first_name last_name

to the LISTSERVer address.

Sincerely,

dr ir Herman J. Woltring, M.Sc.
<ugdist@hnykun53.bitnet, ugdist@nici.kun.nl, na.woltring@na-net.ornl.gov>

Research associate, Nijmegen Institute for Cognition Research and
   Information Technology, Nijmegen University, The Netherlands
Partner, CAMARC-project  &  moderator, Biomch-L@hearn.bitnet

Brussellaan 29,  NL-5628 TB  EINDHOVEN,  The Netherlands
Tel. (private) +31.40.480869, voice/fax/modem +31.40.413744

CAMARC (Computer Aided Movement Analysis in a Rehabilitation Context) is
a project under the Advanced Informatics in Medicine action (AIM) of the
Commission of the European Communities (CEC), with academic, industrial,
public-health, and independent partners from Italy, France, U.K. and The
Netherlands.  Its scope is pre-normative and precompetitive.

Biomch-L (Biomechanics and Movement Sciences) is an electronic bulletin
operated under LISTSERV@HEARN.BITNET, with more than 325 subscribers world-
wide on EARN/BITNET/NETNORTH, the Internet, UUCP/EUnet, JANET, and various
other interrelated networks.  To subscribe, send the one-line request
                SUBSCRIBE BIOMCH-L First_name Last_name
to LISTSERV@HEARN.BITNET (or to LISTSERV@NIC.SURFNET.NL).

------------------------------

Date: Thu, 22 Aug 1991 09:15 +8:00
From: "OZais@cutmcvax.cs.curtin.edu.au"@cc.curtin.edu.au
Organization: Curtin University of Technology, Computing Science
Subject: CFP: AI, Simulation and Planning '92

                          **** CALL FOR PAPERS ****
                                   AIS'92

                          3rd Annual Conference on
                ARTIFICIAL INTELLIGENCE, SIMULATION & PLANNING
                                    in
                           High Autonomy Systems

                               sponsored by
        Schools of Computing Science and Electrical & Computer Engineering
                       Curtin University of Technology

                              Conference theme:

                  INTEGRATING PERCEPTION, PLANNING AND ACTION
                  8 - 10 July, 1992, Perth, Western Australia

The design and development of autonomous systems requires expertise from a
multitude of areas.  The theme of this conference is  the integration of
perception, planning and action using a task-directed approach.  Years of
research in fields of vision and perception, planning, knowledge
representation, modelling, reasoning, control, simulation and many others,
have culminated in the development of models that are useful in limited and
separate domains.  However, integration of these modules in automated
intelligent systems has proved too computationally expensive.  In all these
fields there is a shift towards situated reasoning and task directed problem
solving.  Action requires intelligent decision making, and the choice of
action is based on results of close interaction between task directed sensory
perception and situated reasoning.  The underlying issue is the
representation and modelling of the relevant aspects of the environment, and
problems encountered here are closely related to those in intelligent
simulation.  Techniques in computer simulation can provide useful models that
serve as the foundation for simulating the behaviour of autonomous agents,
and such simulation can provide meaningful insight into the agent's
behaviour. This conference seeks to provide an opportunity for researchers
from a range of disciplines and application domain projects to interact and
address these issues.


SUGGESTED TOPICS:
The topics covered in this conference will include (but will not be limited to)
the conceptual as well as application-oriented issues in the following areas:

* Active vision systems with task-oriented or purposive problem-solving 
* Temporal, spatial, and situated reasoning in automated planning
* Integration of qualitative and quantitative reasoning
* Process planning and control
* Reasoning with uncertainty
* Rule-based modelling and simulation
* Intelligent integrated systems
* Learning in highly autonomous systems
* Adaptive and neural based control
* Fuzzy logic control systems
* Distributed and parallel intelligent simulation

INVITED SPEAKERS:
Rodney A. Brooks     AI Robotics laboratory, MIT
Bernard Zeigler      University of Arizona
Kit Po Wong          University of Western Australia

CONFERENCE CO-CHAIRS:
Dorota Kieronska                       Svetha Venkatesh
School of Computing Science            School of Computing Science
Curtin University of Technology        Curtin University of Technology
Western Australia                      Western Australia
dorota@cutmcvax.cs.curtin.edu.au       svetha@cutmcvax.cs.curtin.edu.au

ORGANIZING COMMITTEE:
Paul A. Fishwick       University of Florida, USA
Franz Pichler          University of Linz, Austria
Jerzy W. Rozenblit     University of Arizona, USA
Suleyman Sevinc        University of Sydney, Australia
Robin Stanton          Australian National University, Australia

PROGRAMME COMMITTEE:
Narendra Ahuja         University of Illinois, USA
Panos Antsaklis        Notre Dame University, USA
E. Balagurusamy        Institute of Public Enterprise, India
Felix Bretschneider    Siemens, Germany
Terri Caelli           University of Melbourne, Australia
F. Y. L. Chin          University of Hong Kong, Hong Kong
Silvano Colombano      NASA Ames Research Centre, USA
John Debenham          University of Technology Sydney, Australia
Vu Duong               Schlumberger Laboratory, France
Norman Foo             University of Sydney, Australia
Michael Georgeff       Australian AI Institute, Australia
Roderick Girle         Griffith University, Australia
Witold Jacak           Technical University of Wroclaw, Poland
Tag Gon Kim            University of Kansas, USA
C.T. Leondes           University of Washington, USA
Olivier Monga          INRIA, France
Ram Nevatia            University of Southern California, USA
Herbert Praehofer      University of Linz, Austria
Ethan Scarl            Boeing Computer Services, USA
Reid Simmons           Carnegie Mellon University, USA
Chuck Thorpe           Carnegie Mellon University, USA
C.P. Tsang             University of Western Australia, Australia
Haruki Ueno            Tokyo Electrical Institute, Japan

PANEL CHAIRS:
Paul A. Fishwick       University of Florida
Jerzy W. Rozenblit     University of Arizona

LOCAL ARRANGEMENTS:
Mary Simpson
Ken Swain
Lance C. C. Fung
Wladyslaw Mielczarski

SUBMISSIONS:
Five (5) copies of extended abstracts (2000 words) must be received by
November 1, 1991 (hard copies only)  A separate sheet should contain the
title of the paper, full names, addresses and email-addresses of the authors,
a short (200 words) summary and keywords indicative of the theme of the
paper.  The abstracts will be refereed, and the authors will be notified of
acceptance by January 15, 1992.  Camera-ready copy sheets will be sent to
authors of accepted papers, and the final papers will be due on March 2,
1992.  The final proceedings will be published by IEEE Press. After the
conference, selected authors may be requested to submit a paper in special
issues of archival journals related to the conference theme.
One author of each paper is expected to present the paper at the
conference.  Any technical inquiries should be directed to one of the
co-chairs of the conference.  All submissions and registration queries should
be directed to:

             Ms. Mary Simpson
             AIS'92
             School of Computing Science
             Curtin University
             PO Box U1987
             Perth, WA 6001
             Australia

             Phone:     +61 - 9 - 351 7298
             Fax:       +61 - 9 - 351 2819
             email:  OZais@cutmcvax.cs.curtin.edu.au


TIMETABLE:
Extended abstracts due:              1 November 1991
Notification of acceptance:          15 January 1992
Formatted papers due:                   2 March 1992
Registration for one author:            2 March 1992
Conference dates:                     8-10 July 1992

------------------------------

Date: Sat, 24 Aug 91 00:34:02 GMT
From: dfuller@ADS.COM (Donna Fuller)
Organization: Advanced Decision Systems, Mtn. View, CA (415) 960-7300
Subject: Sr. Image Understanding Scientist/Engineer Position

Advanced Decision Systems is an R&D and Systems development company
involved in government research and commercial systems.  ADS specializes
in designing and building advanced technology solutions for complex problems.
We currently have an opening for a Senior Image Understanding Scientist/
Engineer.

Job Description/Requirements:
Research, design, and development of a large multi-sensor image understanding
system.  Development environment consists of Encore Multimax, TMC Connection
Machine, and Sun workstations in C/C++/C*.  Responsibilities will include
algorithm development, knowledge elicitation, and task management.

MS/PhD and experience developing image understanding systems preferred.  
US citizenship is required.

Please send your resume to:
Advanced Decision Systems
Dept. IUS
1500 Plymouth Street
Mountain View, CA 94043

------------------------------

End of VISION-LIST digest 10.36
************************
