Copy Link
Add to Bookmark
Report
VISION-LIST Digest Volume 12 Issue 13
VISION-LIST Digest Thu Mar 18 13:23:51 PDT 93 Volume 12 : Issue 13
- ***** The Vision List has changed hosts to TELEOS.COM *****
- Send submissions to Vision-List@TELEOS.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@TELEOS.COM
- Access Vision List Archives via anonymous ftp to FTP.TELEOS.COM
Today's Topics:
Announcing Object Recognition Toolkit (ORT) version 2.3
Matrox MVP-AT Card needed!
Stereo algorithms, looking for
New job
Graduate Studentship: Computer Aided Determination of High Risk Mammograms
Position in Image Compression at HP Labs
ICARCV'94 call-4-papers
------------------------------
Date: Mon, 15 Mar 93 19:50 GMT
From: The Maverick <ATAE@spva.physics.imperial.ac.uk>
Subject: Announcing Object Recognition Toolkit (ORT) version 2.3
[ This code may be found in the Vision List Archive in the
SHAREWARE subdirectory. phil... ]
Object Recognition Toolkit (ORT) Version 2.3
============================================
ORT is a collection of pixel grouping S/W in C aimed for use on Unix platforms
(tested on Sun4, Decstation, Iris, X86 under Linux etc..). It is available
as a compressed tarfile (ORT-2.3.tar.Z) via anonymous ftp from ftp.teleos.com
[131.119.250.108] in the directory VISION-LIST-ARCHIVE/PUBLIC. Currently the
starting point for ORT is an edge detected image. The S/W is in the form of
filters and includes a displayer for use on COLOUR workstations under X11R4/5.
All the S/W comes under the GNU general public licence. Also included are
LaTeX copies of papers on some of the S/W (FEX, LPEG). Basically ORT allows
you to create progressively more complex groupings of a set of edge pixels
[lines/arcs ->parallel/collinear/junctions -> triplets/corners/polygons].
BTW, if you have a LaTeX style file for PAMI or any other AI/IP/CV journals
I'd be most grateful for a copy. Changes from previous versions are:
Version 2.1:
1. All programs are now default driven
2. PGM format as standard with option for raw images
3. Added new options to LPEG and IPEG
4. IPEG is fully general and very much faster
5. LPEG now has option to consider pairs of collinear lines as real lines
Version 2.3:
1. Changed names of programs and converted to lowercase characters
2. Added support for rectangular images
3. Fixed problem with HUGE value being too big for some compilers (eg
gcc on 386 and 486PCs.)
4. Now all programs (except ort2image) may be chained using Unix pipes
5. The ID number assigned to Lines and Circular arcs is constant. Before I
used to change ID numbers when grouping the lines. This caused some
people problems.
6. All programs are faster (specially ipeg).
7. Fixed problem (actually a C bug I think :) in ipeg which made it give
segmentation fault for no reason. This was due the use of bcopy.
If you set the quality factor very low (eg 0.3) it usually it requires
too much memory for average machines and crashes. Since I don't know how
to check for available memory on-the-fly I left this alone.
8. Restructured the directories
The features defined within ORT represent the complete set required to
recognize any polyhedral object (view point independently) starting from edge
information. Available later this year will be an Edge detection module,
and hopefully also a generalised curve grouping system.
CODE DESCRIPTION
Liste List handling library in C by Jean-Paul Schmidt formerly
of University of Surrey, UK. [Version 1.2]
chainpix Pixel Chainning code by Geoff A.W. West and Paul L. Rosin
of Curtin University, Australia [Version 1.2]
fex Segments chained pixel lists produced by chainpix
into straight-line segments and circular arcs. [Version 1.7]
lpeg Low-level straight-line grouping [Version 1.9]. Groups
straight-line segments produced by fex into:
Parallel overlapping Parallel non-overlapping
Collinear V,L,T, and Lambda Junctions
ipeg Intermediate-level grouping [Version 2.1]. Groups sets
produced by lpeg into:
Triplets (barends, Z)
Corners (3 lines sharing a junction point)
Polygons
DisplayGrps X11R4/5 viewer for the above groupings/segments by
Jean-Paul Schmidt and Ata Etemadi [Version 1.2]
All contributions/comments to the distribution are most welcome. If you
have any problems I'll be glad to help.
best regards
Ata <(|)>.
| Mail Dr Ata Etemadi, Blackett Laboratory, |
| Space and Atmospheric Physics Group, |
| Imperial College of Science, Technology, and Medicine, |
| Prince Consort Road, London SW7 2BZ, ENGLAND |
| Internet/Arpanet/Earn/Bitnet atae@spva.ph.ic.ac.uk or ata@c.mssl.ucl.ac.uk |
| Span SPVA::atae or MSSLC:atae |
| UUCP/Usenet atae%spva.ph.ic@nsfnet-relay.ac.uk |
------------------------------
Date: Tue, 16 Mar 93
From: Herbert Hufnagl <HU@flexaut.tuwien.ac.at>
Subject: Matrox MVP-AT Card needed!
Our Institute is in urgent need for a Matrox MVP-AT Card. Because
this card is not of the latest technology, we hope to find
somebody who has a used MVP-Card that is not in use anymore, and who
would sell it cheap. We do not need the card for commercial use - we
use it to teach students at the University of Technology, Vienna.
Thank you, Herbert
Herbert Hufnagl
Institute for Flexible Automation
University of Technology Vienna
Gusshauustr. 27-29/361
A-1040 Vienna/AUSTRIA
Tel.: ++43/1/58801/3681
Fax.: ++43/1/5055983
------------------------------
Date: Wed, 17 Mar 1993 16:56:24 -0500
From: Arun Katkere <katkere@eecs.umich.edu>
Subject: Stereo algorithms, looking for
I am looking for some stereo algorithms to create test data for some modules
of an outdoor navigation system we are building. If you have a stereo
algorithm that you'd like to share, please let me know.
We have MaxVideo20 available. So, datacube implementations of stereo
algorithms will be appreciated. But, for my immediate purposes, software
implementations are sufficient.
I'd appreciate some pointers/paper suggestions on creating dense range maps
and on integrating multiple range maps. Some notes about actual experience in
using stereo in navigation tasks (like low level obstacle avoidance) will be
a bonus.
Send responses to "katkere@eecs.umich.edu" and if there is sufficient
interest (and I get some responses!) I'll post a summary.
Thanks,
arun
Arun Katkere | Artificial Intelligence Laboratory
katkere@eecs.umich.edu | The University of Michigan
O:(313)763-1563 | R:(313)761-9462 | 147 ATL, 1101 Beal Avenue
fax: (313)763-1260 | Ann Arbor, MI 48109-2110
------------------------------
Date: Thu, 18 Mar 1993 15:10:21 GMT
From: slack@starbase.mitre.org (M. G. Slack)
Organization: The MITRE Corporation
Subject: new job
The MITRE Corporation in McLean Virginia has an immediate need to fill a
mid-level position in applied robotic intelligence/vision.
The position is for a Ph.D. in CS or a masters with 3 to 5 years experience in
developing active vision systems to support the autonomous operation of robots.
We would prefer a person who not only has a strong background in active vision,
the ideal candidate would have a working knowledge of the fields of Planning,
and Reactivity as they relate to autonomous robots. The person should also
have strong programming skills in both C and Lisp.
The person will join the team of the Autonomous Systems Laboratory (ASL). In
particular the person will be working with the PRISM stereo system developed by
Teleos in order to develop visual routines which will be used to provide
autonomous robots with perceptual information about the world.
In addition to robotic intelligence, MITREUs Artificial Intelligence Technical
Center is involved in a number of other AI related areas, including planning,
heterogeneous databases, natural language processing, intelligent terrain
evaluation, decision-support systems, and information fusion.
MITRE is a not-for-profit Corporation chartered by the Federal Government for
work in the public interest. When you work for MITRE you MUST be a US. citizen
and you can't have any outside jobs except teaching positions. Salary is
determined on an individual basis according to degree level and years of
experience since the B.S. degree. The retirement and benefits are competitive
with most defense contractors.
If you are interested in this position please contact me ASAP (via email).
Marc Slack, Ph.D.
Lead Scientist
Mail Stop Z401
The Mitre Corporation
7525 Colshire Drive
McLean, Virginia 22140-3481
slack@starbase.mitre.org
----------------------------------------------------------------------
Date: 16 Mar 1993 14:20:10 -0000
From: pt@curie.lif.icnet.uk (Paul Taylor <pt@curie.lif.icnet.uk>)
Organization: Imperial Cancer Research Fund
Subject: Graduate Studentship: Computer Aided Determination of High Risk Mammograms
COMPUTER AIDED DETERMINATION OF HIGH RISK MAMMOGRAMs
Mammographic dysplasia (Wolfe grading) is a well established although
controversial risk factor for breast cancer (Boyd et al, 1984). With the
advent of the national screening programme most women will now have regular
mammograms and in addition to detecting small cancers they could also be used
to determine which women are at increased risk of cancer. This has two
potential benefits: tailoring the screening interval to match the risk of
disease, and determining which women might be suitable for breast cancer
prevention studies, such as the currently proposed tamoxifen prophylaxis study
(Cuzick, 1992).
Determining the extent of mammographic dysplasia is subjective and requires
training which few British radiologists have obtained. However, the problem
is basically a texture recognition problem and should be amenable to automatic
determination by computer (Caldwell et al, 1990).
The project consists of developing this approach and will be supported by the
ICRF computing group which has an active interest in image analysis, the ICRF
breast cancer unit at Guy's Hospital which will supply mammograms and clinical
advice, and Dr. Huw Gravelle, who is Britain's leading expert on Wolfe
grading of mammograms. The work will consist of developing and implementing
statistical methods for texture analysis and a dedicated SUN workstation and
appropriate software for image analysis will be available. A strong
background in statistics and computing is required, as well as an inquisitive
mind and a desire to apply this knowledge to medical problems.
Reply to:
Dr. Jack Cusick
Imperial Cancer Research Fund
61 Lincoln's Inn Fields
LONDON WC2A 3PX
ENGLAND
tel: 071 269 3007
email for Jack can be sent care of): pt@acl.lif.icnet.ac.uk
References:
Boyd NF, O'Sullivan B, Fishell E, Simor I, Cooke G (1984). Mammographic
patterns and breast cancer risk: Methodologic standards and contradictory
results. J Natl Cancer Inst; 72: 1253-59.
Cuzick J (1992). Methodologic issues in the chemoprevention of breast cancer.
Cancer Detection & Prevention; 16: 81-86.
Caldwell CB, Stapleton SJ, Holdsworth DW, Jong RA, Weiser WJ, Cooke G & Yaffe
MJ. Characterisation of mammographic parenchymal pattern by fractal
dimension. Phys Med Biol; 35: 235-247.
------------------------------
Date: Mon, 15 Mar 1993 21:09:07 GMT
From: drukarev@hplabsz.hpl.hp.com (Alexander Drukarev)
Organization: Hewlett-Packard Laboratories Palo Alto,CA
Subject: Position in Image Compression at HP Labs
HEWLETT-PACKARD LABORATORIES has an employment opportunity in
IMAGE COMPRESSION
We are looking for an individual who will play a leadership role in
research and development of image compression and processing algorithms.
The main responsibilities will include: identification of the research
directions in support of the current and new company businesses, developing
advanced compression algorithms, consulting in the area of image processing
and compression.
Required experience includes: PhD in CS/EE or related field, minimum 2 years
of related experience. Experience in video processing and compression is a
plus.
HEWLETT-PACKARD LABORATORIES offers a stimulating work environment,
competitive salaries and an excellent benefit package. We are an equal
opportunity employer. Please, send your resume to: Alex Drukarev,
Hewlett-Packard Labs, 1501 Page Mill Rd., Bldg. 3U, Palo Alto, CA 94304.
------------------------------
Date: Thu, 18 Mar 93 13:36:15 +0800
From: hw@ntu.ac.sg (Wang Han)
Subject: ICARCV'94 call-4-papers
CALL FOR PAPERS
ICARCV'94
THIRD INTERNATIONAL CONFERENCE ON
AUTOMATION, ROBOTICS AND COMPUTER VISION
8-11 NOVEMBER 1994
The Third International Conference on Automation, Robotics and Computer
Vision, ICARCV'94, will be held in Singapore on 8-11 November 1994. The theme
of ICARCV'94 is "Intelligent Automation". The conference will provide a forum
for automation professionals, manufacturing engineers and academic researchers
to exchange up-to-date technical knowledge and experience. The conference
will focus on both theory and applications. In addition to the technical
sessions, there will be plenary, invited and tutorial sessions. Also an
exhibition will be held in conjunction with the technical and poster sessions.
Papers describing original work in, but not limited to, the following
research areas are invited:
*ROBOTICS* *COMPUTER VISION*
Robot Control Image Processing and Interpretation
Mobile Robots and Navigation 3-D/Colour/Stereo Image Analysis
Task Planning Dynamic Scene Analysis
Intelligent Sensors and Actuators Vision Systems
Micro-Robots Pattern Recognition & Applications
Robot Design and Simulation Machine Intelligence
Multiple Robots
*INTELLIGENT AUTOMATION*
*CONTROL THEORY AND APPLICATIONS*
Instrumentation Systems *SIGNAL PROCESSING AND APPLICATIONS*
Flexible Manufacturing Systems *AI & EXPERT SYSTEMS*
Process Automation & Networking *NEURAL NETWORKS AND FUZZY SYSTEMS*
Man-machine Interaction *REAL TIME SYSTEMS*
On-line Process Monitoring&Control *PETRI NETS AND APPLICATIONS*
Factory Modelling and Automation
Submission Procedures :
Papers must be written in English and should describe the work in detail.
Three copies of the full paper must be submitted for review. Please include in
the forwarding letter the author's name to whom the correspondence should be
addressed, address for communication, contact telephone and fax numbers, e-
mail addresses and also the broad classification under which the paper can be
included. Please indicate your preference for regular or poster session.
Participants will be required to register upon notification of acceptance
of their papers. Papers will be published in the conference proceedings only
if at least one of the authors is officially registered. The length of the
final paper in the proceedings will be limited to a max of 5 pages (5 mats B4
size), including figures, tables and references. There will be a charge of
S$100 per page for papers exceeding 5 pages.
SELECTED PAPERS WILL BE PUBLISHED IN INTERNATIONAL ENGINEERING JOURNALS.
Invited Sessions : The Programme Committee is also soliciting proposals for
invited sessions focusing on new or emerging technologies relevant to the
ICARCV'94 theme. Please submit proposals to the Technical Programme Chairman,
Professor N. Sundararajan, before 31 March 1994.
Exhibitions : Exhibition of products related to the ICARCV'94 theme is
invited. Information is available upon request.
Please submit three copies of the full paper for review to :
Professor N. Sundararajan
c/o ICARCV '94 Conference Secretariat
Institution of Engineers, Singapore
70 Bukit Tinggi Road, Singapore 1128
Republic of Singapore
Tel : (65) 469 5000, Fax : (65) 467 1108
Telex : RS 22992 IESIN
E-mail: ensundara@ntuvax.ntu.ac.sg (Internet)
ensundara@ntuvax.bitnet (Bitnet)
AUTHOR'S SCHEDULE :
30 April 1994 Receipt of full paper
31 May 1994 Notification of Acceptance
31 July 1994 Receipt of Camera Ready Manuscript
31 August 1994 Advance Registration Deadline
ORGANISERS:
School of EEE, Nanyang Technological University
Institution of Engineers, Singapore (IES)
CO-SPONSORS:
Defence Science Organisation(DSO)
GINTIC Institute of Manufacturing Technology(GIMT)
Instr. & Control Society, Singapore
National Computer Board(NCB)
National Science and Technology Board(NSTB)
Singapore Computer Society(SCS)
Singapore Industrial Automation Assoc.
IN COOPERATION WITH:
IEEE Robotics and Automation Society
IEEE Neural Network Council
------------------------------
End of VISION-LIST digest 12.13
************************
