Copy Link
Add to Bookmark
Report
VISION-LIST Digest Volume 13 Issue 37
VISION-LIST Digest Wed Aug 24 15:40:16 PDT 94 Volume 13 : Issue 37
- ***** The Vision List host is 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:
WWW vision paper reference archive CORRECTION
Computer Vision Home Page (reminder)
Pan-tilt unit
ACR-NEMA Source Code
Occupancy question ...
Point in a rectangle
Suggestions for digital cameras
Req. for info: real-time image proc./recog. hardware...
"Steady-cam" vision chip desired
Request : Edge Linking Algorithm Information
Scale DTs
(GL) perspective views & camera's focal length
Re: Matching translated tesselation
Seeking for geometric hashing code and data
Range coding
S: guest scientist
M.Sc. in Knowledge Engineering
Wanted information about Carleton U.
----------------------------------------------------------------------
Date: Wed, 3 Aug 1994 10:57:51 +0100
Subject: WWW vision paper reference archive CORRECTION
From: Andrew Smallbone <A.Smallbone@qmw.ac.uk>
Sorry an error appeared in the URL's for the WWW Article Archive
announced in the VISION-LIST digest 13.34
WWW Vision and Image Processing Article Archive
==============================================
As part of an awareness project on high performance computing, the
``London and South East Centre for High Performance Computing'' has
created a WWW vision and image processing archive (there are also high
performance computing, functional programming, neural networks, and
computational mathematics archives). _You_ can add your own articles
to the archive using a series of forms. The archive is available at
URL(s):
monochrome users:
http://www.lpac.qmw.ac.uk/bin/monoFilter?/SEL-HPC/Articles/index.html
Kitsch colour page:
http://www.lpac.qmw.ac.uk/SEL-HPC/Articles/index.html
For more information see VISION-LIST digest 13.34 or access above pages
Sorry for any inconvenience
------------------------------
Date: Mon, 22 Aug 94 21:06:44 EDT
From: Mark.Maimone@A.GP.CS.CMU.EDU
Subject: Computer Vision Home Page (reminder)
COMPUTER VISION HOME PAGE
REMINDER POST
The Computer Vision Home Page is a collection of vision-related resources
available on the Internet. It is written in the form of a World Wide Web
(WWW) Hypertext document and contains pointers to more than 4000 test
images, FTP and GOPHER archives, publications, source code, newsgroups,
conferences, and more than 40 research groups' home pages.
To access the Computer Vision Home Page, just run a program that knows about
the World Wide Web (e.g., NCSA Mosaic) and open one of these URLs:
http://www.cs.cmu.edu:8001/afs/cs/project/cil/www/vision.html
or http://www.cs.cmu.edu:8001/afs/cs/project/cil/www/txtvision.html
(the first one includes lots of useful icons but is slow to load, so you
might prefer the second which is text-only)
I'd appreciate hearing any suggestions you might have about how to improve
the page (e.g., anything important I've left out, something that doesn't
work for more than one day, something you found really useful). But at this
point I can't offer to maintain it in any other format. :-(
ACCESSING THE WEB
Anyone with a telnet-capable Internet connection can access the World Wide
Web. If you have a Unix workstation, Macintosh, or PC with Windows you
might like to try the Mosaic program, which is available by anonymous FTP to
ftp.ncsa.uiuc.edu, cd Mosaic. This is a neat program that supports text,
hypertext, images, animations and sounds (if your system is already
configured for them). However, if you're limited to a text-only interface
(but can still telnet) there are still some text-only browsers out there you
can try (but it's not nearly as much fun!):
If you're near: Telnet to:
Switzerland: info.cern.ch or 128.141.201.74
Kansas, USA: ukanaix.cc.ukans.edu (login as www)
New Jersey, USA: www.njit.edu (login as www)
Israel: vms.huji.ac.il or 128.139.4.3 (login as www)
Slovakia: sun.uakom.cs (slow link; use only from nearby)
Hungary: fserv.kfki.hu (slow link, login as www)
Finland: info.funet.fi or 128.214.6.100
Finally, if you are not directly connected to the Internet, as a last resort
you can send an email message to listserv@info.cern.ch with the body:
send http://www.cs.cmu.edu:8001/afs/cs/project/cil/ftp/html/txtvision.html
You'll get a very poorly formatted copy of the vision page, but one which
has all the same information. BTW, this is an experimental service, and
only delivers then first 1000 or so lines of text files; no images or
archive files.
Mark Maimone phone: +1 (412) 268 - 7698
Carnegie Mellon Computer Science email: mwm@cmu.edu
WWW: http://www.cs.cmu.edu:8001/afs/cs/usr/mwm/ftp/www/home.html
------------------------------
Date: Mon, 15 Aug 94 15:40:21 -0700
From: dgaw@coactive.com (David Gaw)
Subject: Pan-tilt unit
There have been several inquiries in this group about pan-tilt units
for vision and robotics. You may want to take a look at a commercially
available computer-controlled pan-tilt tracking mount that is
manufactured by Directed Perception, Inc. Their PTU-46-17.5 pan-tilt
unit has been designed to be low-cost ($1800 in the US/Canada, $2100
international), fast (e.g., over 300 degrees/sec), and accurate (e.g.,
resolution of 3.09 seconds/arc per position). Below is an overview:
Information can be obtained from:
Directed Perception, Inc.
1451 Capuchino Avenue
Burlingame, CA 94010-3308
(415) 342-9399, FAX: (415) 342-9199
GENERAL FEATURES OF THE PTU-46-17.5 PAN-TILT TRACKING MOUNT
Maximum Rated Payload: over 4 lbs
Maximum Speed: over 300 deg/second
Resolution: 3.086 arc minutes
Tilt Range (approx): minimum 31 deg up and 47 deg down (78 deg range) with
option of 80 deg down (111 deg range)
Pan Range (approx): +-159 deg (318 deg range) with
option of +-170 deg (340 deg range)
* Simple to use from any RS-232 terminal or computer
* Precise control of position, speed & acceleration
* Small form factor
* Allows on-the-fly position and speed changes
* Self calibration upon reset
* Power consumption can be controlled from host
* ASCII command mode for simplicity, binary commands available for
efficient program control
* Constant current bipolar motor drives for increased performance
and control
* Wide range of allowable DC power input from an unregulated source
AVAILABLE OPTIONS
* AC/DC Power Supply (model PTU-PS)
* Higher resolution and payload capacity
* Nodal version
------------------------------
Date: Thu, 11 Aug 94 10:00:17 -0400
From: perry@enet.hq.ensco.com (John L. Perry)
Subject: ACR-NEMA Source Code
Does anyone have source code (Mac or other platforms) for reading/writing
ACR-NEMA formats, especially the new DICOM standard (PS3) ? Tks!
John L. Perry
Director, Advanced Technology
ENSCO, Inc.
5400 Port Royal Road
Springfield, Virginia 22151
U.S.A.
Phone #: (703) 321-4540
Fax #: (703) 321-7759
Email (Internet): perry@enet.hq.ensco.com
------------------------------
Date: 3 Aug 1994 19:05:31 GMT
From: mookherj@fiu.edu (Jay Mookherje)
Organization: Florida International Univeristy
Subject: Occupancy question ...
A rectangular plane is subdivided into N rows and N columns, which subdivides
the whole area into N*N small ZONE's. Now, if I drop any arbitrary rectangle
onto this plane, how would I find out which ZONE's will it occupy ? Occupancy
can be partial or complete.
Please forward all responses to mookherj@fiu.edu.
Thanks in advance,
Jay
------------------------------
Date: 4 Aug 1994 07:00:00 GMT
From: mookherj@fiu.edu (Jay Mookherje)
Organization: Florida International Univeristy
Subject: Point in a rectangle
Maybe this question has been asked a million times before, but I would
really appreciate if someone could tell me an efficient way of finding if
a particular point lies inside a rectangle (of arbitrary orientation).
Please direct all responses to mookherj@fiu.edu.
Thanks in advance,
Jay
------------------------------
Date: Tue, 2 Aug 94 16:17:31 EDT
From: kumar@sarnoff.com (Rakesh Kumar x2832)
Subject: Suggestions for digital cameras
Hi
I am soliciting recommendations for digital b/w or color cameras which
can be directly connected to a MAC or SUN sparc station. We are interested
in purchasing cameras which cost less that $ 10,000/- or so. If you have used
these cameras, I would be interested to know of your experience in terms
of quality and resolution of images, ease of operation and finally
what software and hardware accesories need to go with the digital camera ?
Thanks
Rakesh (Teddy) Kumar
David Sarnoff Research Center Phone: 609-734-2832
CN5300, Princeton, NJ-08543 EMAIL: kumar@sarnoff.com
------------------------------
Date: 8 Aug 1994 06:16:30 GMT
From: sergeb@cco.caltech.edu (Serge J. Belongie)
Organization: California Institute of Technology, Pasadena
Subject: req. for info: real-time image proc./recog. hardware...
Hello,
I would like to get advice from people regarding the following:
I am interested in making a purchase of a real-/quasireal-time
image processing system that is capable of the following:
- grabbing frames at at least 5 or 6 Hz
- displaying the frames as they come in
- convolving input images with 8x8 kernels
- basic signal processing, such as image multiplication and addition,
subsampling, etc.
- extensibility/inclusion of recognition/template-matching/feature-
tracking hardware
The application I have in mind is real-time pyramid formation
(Gaussian & Laplacian) for feature extraction and recognition.
We have looked into the systems Datacube offer; they seem very
nice but are pricey (~$50,000). I'm curious what options are out
there for that much money in terms of Sparc-based/C-friendly
image processing environments.
Any info will be appreciated!
Serge Belongie
Caltech EE dept
------------------------------
Date: Mon, 8 Aug 1994 07:49:33 -0400
From: "Ted A. Camus" <tac@cs.brown.edu>
Subject: "steady-cam" vision chip desired
I am interested in a small chip to perform real-time image stabilization,
similar to the "steady-cam" feature of modern camcorders. (The latter in
fact would be ideal, if commercially available). This would be used to
stabilize the lateral movement a small robot "helicopter" in windy
conditions. Advthanksance for any replies --
-- Ted --
------------------------------
Date: Wed, 24 Aug 94 09:30:03 -0400
Subject: Request : Edge Linking Algorithm Information
From: sjpratt@cat.syr.edu
I am presently working on a method of linking edges from a binary edge map
created by an edge detection program (ie. Canny, Sobel, etc.). I am looking
for an established edge linking algorithm to which I can compare my results.
I would appreciate any comments, good or bad, about any edge linking method
that you might have used or learned about. What is the accepted edge linking
method in use today? I would ideally like to find some C code implementations
of these algorithms, if possible.
Please email me your opinions on these methods and if there is enough interest,
I will compile a summary of algorithms, comments and hopefully, coded
implementations.
Thank you,
Steve Pratt
sjpratt@cat.syr.edu
------------------------------
Date: Fri, 5 Aug 1994 17:30:19 +0800 (WST)
From: geoff@cs.curtin.edu.au
Subject: scale DTs
Hi,
I have recently come across a type of distance transform called the
"scale distance transform". Can anyone enlighten me on what this is
and how it is different from the normal DT?
Thanks in advance,
Geoff West,
School of Computing,
Curtin University,
Perth, Western Australia.
Email: geoff@cs.curtin.edu.au | ... __o
Fax: + 61 9 351 2819 |... -\<,
Tel: + 61 9 351 3005 |....(_)/(_)
------------------------------
Date: 9 Aug 94 02:16:34 GMT
From: shahidi@caip.rutgers.edu (Ramin Shahidi)
Organization: Rutgers Univ., New Brunswick, N.J.
Subject: (GL) perspective views & camera's focal length
Hello
Does any body have any idea how to relate the perspective
characteristics for simulating a camera view to the focal length of
the camera you are simulating.
Basicly I am simulating camera views using GL and I would like to
figure out the focal length of the camera I am simulating.
I have the distance from th object to camera, my camera's viewing
angle, the image size... clipping planes ...
Has anybody figured out these relationships..
All refernces/advices and suuggestions are greatly appreciated.
please reply by email
shahidi@vizlab.rutgers.edu
**ramin
------------------------------
Date: 3 Aug 1994 18:56:07 GMT
From: oconnn@rpi.edu (Nathan John O'Connor)
Organization: Rensselaer Polytechnic Institute, Troy, NY.
Subject: Re: Matching translated tesselation
Keywords: Phase correlation
In response to :
/******************************************************************************/
Hi,
Can anyone give me some tips/ideas on how to solve the following problem
I have 2 grids (256x256). Each grid has 40 locations scattered all over it.
The 2 "patterns" of locations on each grid is precisely the same except that
the one grid has an offset in both the x and y directions. The offsets are
unknown. Some of the locations in each "pattern" may be missing, but not
more than say 30% of the 40 locations.
I need an algorithm to determine the x and y offsets. I have one solution,
but it takes much too long on my 386DX40. It is basically an optimizing
algorithm that determines the position where the one pattern multiplied with
the other pattern gives a maximum value.
email : s9151893@mella.ee.up.ac.za
/*****************************************************************************/
If you can perform 2-D ffts fast enough one routine you might want to consider
is the Phase Correlation algorithm by Castro et. al. It is pretty easy to implement.
It can be non-iterative for non-rotated patterns. It determines the translation
vector (dx,dy).
Good Luck,
Nathan
email :
oconnor@ipl.rpi.edu
------------------------------
Date: 24 Aug 1994 01:24:59 -0700
From: separk@pollux.usc.edu (SEONGBIN PARK)
Organization: University of Southern California, Los Angeles, CA
Subject: seeking for geometric hashing code and data
Hello,
Is there any ftp site where I can download the code for geometric hashing and
sample data ? Any information would be really appreciated !
Regards,
Seongbin Park (separk@pollux.usc.edu)
------------------------------
Date: Thu, 4 Aug 1994 16:26:52 GMT
From: ronald@kohnolab.dnj.ynu.ac.jp (Wang KEFEI / M1 )
Organization: Division of Electrical and Computer Engineering, Yokohama
National University, Japan.
Subject: Range coding
Dear sirs,
I am intersted in the research of compression coding for range data.
Dose anyone know where can I get the range data and do you have any
idea about the research of compression coding for range data ?
Please e-mail me if you would like.
Ronald
(e-mail: ronald@kohnolab.dnj.ynu.ac.jp)
------------------------------
Date: 17 Aug 1994 02:06:17 -0700
From: gentner@chaph.usc.edu (Steven Gentner)
Organization: University of Southern California, Los Angeles, CA
Subject: Help Wanted!
Wanted: Student with training in probability theory and computer expertise to
collaborate or consult (paid) with Psychology professor in a project about
perception of symmetry.
Topic: Determining the conditions under which mirror symmetry in random dot
patterns can be perceived. The general procedure in this area of research
is to create a random dot array in one half of the field and then mirror
it to the other half. It is assumed (or claimed) that the resulting
visual stimulus does not have any luminance differences across the axis
of symmetry, and that detection of the axis depends on an ability to see
paired distances among the points.
I believe that there is a flaw in this procedure, and that local brightness
filters of the visual system will detect an unusually high incidence of
clusters which provide the basis for seeing a border.
The research project requires the construction and printing of arrays
having specified probabilities of points, line-by-line across the array,
followed by scans which establish the frequency of point/nonpoint clusters.
Respond to: egreene@bcf.usc.edu
Steven Gentner | "the crux of sorcery is the internal dialog;
Rosemead CA | that is the key to everything. When a warrior
gentner@phakt.usc.edu | learns to stop it, everything becomes possible;"
Don Juan
------------------------------
Date: Wed, 24 Aug 94 09:34:00 +0000
From: Christoph Klauck <klauck@Informatik.Uni-Bremen.DE>
Subject: S: guest scientist
The project IRIS (Image Retrieval for Information Systems) is running at
the Artificial Intelligence group (Prof. Dr. O. Herzog) at the University
of Bremen, Germany. We are looking for a guest scientist
from October 1,1994
to July 31, 1995
Besides research for IRIS we expect teaching activities. The possibility
for an extension of the stay is still open.
The main research activities are expected to be in the area of extracting
feature vectors out of images. These vectors will be used for object
recognition. The images will be classified by their semantical
content that will be used for image retrieval.
Experience in AI techniques and a working knowledge in image
analysis and C/C++ is required. Additional experience in taxonomical
languages and Prolog will be of advantage.
The salary offered will be determined according to the qualifications
and/or experience of the successful applicant.
If qualified please submit a confidential resume as soon as possible to:
Dr. Christoph Klauck
Universitaet Bremen
Arbeitsgruppe Kuenstliche Intelligenz
FB 3 - Informatik
Bibliothekstrasse 1, MZH 5050
D-28359 Bremen
Germany
P.O. Box 33 04 40
D-28334 Bremen
Phone: +49-421 218-4554
-7090 (Secretary: Ms. Ina Chromik)
Fax: +49-421 218-7196
internet: klauck@informatik.uni-bremen.de
Christoph Klauck EMail: klauck@informatik.Uni-Bremen.de
FB 3 - Informatik - AG KI Phone: ++49-421-218-4554 /-7090
Universitaet Bremen Fax: ++49-421-218-7196
Postfach 330 440
D - 28334 Bremen
------------------------------
Date: Wed, 24 Aug 1994 16:46:53 GMT
From: vassilis@westminster.ac.uk (Vassilis Konstantinou)
Organization: University of Westminster
Subject: M.Sc. in Knowledge Engineering
MSc/PgD Diploma in Knowledge Engineering
Full-time (1 year) or Part-time (2 years minimum)
Knowledge Engineering
Knowledge Engineering is the part of computer science concerned
with designing knowledge based computer systems, which exhibit
characteristics we would normally associate with intelligence in
human behaviour.
This course addresses the skills required to develop such
systems for application to real-life problems. It is an advanced
course which aims to to provide students with a thorough working
knowledge of current techniques in artificial intelligence and
knowledge based systems design. The course places particular
emphasis on practical skills which will enable students to
contribute to the development of advanced information technology
within industry and commerce.
Course structure
To obtain the MSc the student must study six taught modules, one
coursework based module and complete a project. This course is
modularised and supports the credit system. Each module is worth
15 credits at level M which are transferable to other
institutions of higher education that use the credit system.
The academic year is divided into two semesters. Full-time
students take three modules in each semester. Part-time students
may take up to four modules per year. In addition, all students
are required to follow 'Knowledge based Systems Design' for two
semesters.
The final component of the course is a project dissertation. The
project is expected to be equal to three months full-time work
and is equivalent to two modules. Full-time students undertake
the project between July and September. It is recommended that
part-time students undertake the project between July and
December at the end of the second year, although a part-time
student may take up to 4 years to complete the MSc course
provided the minimum annual credit requirement, according the
university regulations, is met.
Modules and aims
1) Knowledge Based Systems Design
Introduces current techniques in knowledge representation an
artificial intelligence programming including case studies of
current systems. It runs over two semesters, formal teaching
being given in the first semester, the second semester consisting
of labwork on mini-projects undertaken by students either
individually or as part of a team. Assessment is by coursework
only.
2) Languages for Artificial Intelligence
Introduces 2 major AI programming languages used in industry:
Prolog and Lisp.
3) Logic for Knowledge Representation
Introduces classical logic and it extensions, modal logic,
temporal logic, default and non-monotonic logic. Offering a
theoretical base many practical issues in knowledge engineering.
4) Neural Computation
Introduces the main neural computing architectures and
applications.
5) Computer Vision
Presents computational techniques used in the interpretation of
images and their relation to currently known mechanisms of
biological vision.
6) Machine Learning
Introduces machine learning techniques which enabling a computer
to improve its own performance in the light of experience within
a problem domain.
7) Distributed Artificial Intelligence
Presents architectures and languages used for achieving
co-operative problem solving in a distributed environment will be
described.
8) Natural Language Understanding
Presents and discusses the principles of Natural Language
Understanding and Generation.
9) Uncertain Reasoning Mechanisms
Examines approaches to the handling of uncertain and incomplete
information by knowledge based systems.
Who should apply
The course is open to graduates, from Computer Science and other
technical disciplines, with a first or second class honours
degree from a British University or equivalent qualification.
Each applicant will be considered on his or her own merits. A
flexible attitude will be taken towards entry requirements and it
will be possible to relax the formal educational requirements for
those employed in an industrial context that is related to
Knowledge Engineering.
Information
For further information please contact:
Peter O'Reilly or Vassillis Konstantinou
Artificial Intelligence Division
School of Computer Science
University of Westminster
115 New Cavendish Street
London WIM 8JS
Telephone: (071) 911 5000 Ext. 3602, 3599 or 3598
V. Konstantinou
Head of A.I. Division - School of Computer Science
University of Westminster, 115 New Cavendish Street,London W1M 8JS.
------------------------------
Date: Thu, 11 Aug 1994 12:09:49 GMT
From: alexz@tmx100.elex.co.il (TMX100_2 Alex Zacharov 2396)
Organization: Telrad Ltd.
Subject: Wanted information about Carleton U.
I am thinking about Ph.D. study in Carleton University ( Ottawa, Canada ) and
would like to know about people ( professors, assistants etc. ), who work with
Carleton University in next directions:
- Image processing;
- Image compression;
- Image transmition;
- Image understanding;
- Robot ( technical ) vision;
- Modeling of Human vision;
- etc.
May anyone help me ?
I'll be VERY VERY much obliged for answer.
Please, mail me to alexz@tmx100.elex.co.il
-----------------------
Thanks in advance,
Alexander Zakharov.
------------------------------
End of VISION-LIST digest 13.37
************************
