Subject: Re: Pulsar-LC frame grabber and Windows NT 4.0
From: perry@netcom.com (Perry West)
Date: Thu, 14 Nov 1996 15:34:03 GMT
Hans de Boer (boer@kfih.azr.nl) wrote:
: I am using a Matrox Pulsar LC frame grabber under Windows
: NT 3.51 (Workstation) and am planning to move to Windows NT 4.0.
: However, I have heard that the Pulsar-LC has some real problems
: under NT 4.0. If anyone reading this has experienced such
: problems, can he/she give me some information so I can determine
: if upgrading would be a bad idea for us ?
Hans --
We have used the Pulsar under NT 3.51 and mostly under NT 4.0. Our
challenges have not been so much the operating system but the unusual
camera configuration we need. The main thing to know about using the
Pulsar under 4.0 is to get the appropriate display driver; the version
that works on 3.51 will not work on 4.0. Also, make sure you have the
latest MIL files. The driver and other files are available from Matrox's
BBSs (note that BBSs *is* plural). The files are supposed to be
accessable from Matrox's web site (www.matrox.com), but I couldn't find
them. I couldn't find the BBS numbers, but their local rep should have
them both.
Regards,
Perry West
Subject: Re: Motion estimation
From: Bo Zhang
Date: Thu, 14 Nov 1996 09:57:59 -0500
Guillaume ESCARGUEL wrote:
>
> I'd like to improve the compression ratio of my wavelet transform
> compression algorithm for Video(2 frames/sec). I am looking for hints
> concerning the motion detection, correlation inter or intra frames.
> The main goal is to improve the compression rate without altering the
> quality of the reconstructed video-clip.
>
> Thanks for your help.
>
> My e-mail address : escargue@fairway.ds.cubic.com
>
> Guillaume ESCARGUEL
There is a paper about motion compensation for wavelet transform:
Y.Zhang, S. Zafar,"Motion-compensated Wavelet transform Coding for Color
Vedio Compression", IEEE Trans. on Circuits and systems for Vedio
Technology, vol.2,No. 3, Sep. 1992.
Subject: Re: Ultrasound research?
From: Maurice klein Gebbinck
Date: Thu, 14 Nov 1996 17:43:06 +0100
Tai Luu wrote:
>
> I am currently trying to the ultrasoud reasearch but I can't find any
> resoures on the internet. Could anyone helps me to find it.
You could try the department were I did my graduation research
(yes, it had to do with ultrasound ;-) ):
http://www-kflkg.azn.nl/
Hope this helps,
Maurice
--
Maurice klein Gebbinck
University of Nijmegen, Computing Science Institute
Toernooiveld 1, 6525 ED Nijmegen, The Netherlands
Phone +31 24 3652710, Fax +31 24 3553450
http://www.cs.kun.nl/~mauricek
Subject: Re: Interpolate interlaced video frames
From: Nicolas Devillard
Date: Tue, 12 Nov 1996 13:43:09 +0100
On Tue, 12 Nov 1996, Ingo Elsen wrote:
[...]
> If the scene changes rapidly it may be possible to regard the two fields
> as to be two frames. The aspect ratio changes, so the algorithmns have
> to be adapted. You will have half the frame time then, but also have the
> amount of data to be processed.
Indeed, the frame ratio does not necessarily have to be maintained for
algorithms. Cross-correlation works, for example, if you also interlace
your pattern. Working with fields instead of frames is usually not really
an issue for algorithms, but certainly causes headaches in implementation.
> If you only need one of the original fields and you are not in need of
> the full horizontal resolution, you may use a subsampling scheme of the
> video signal (with half the pixel clock), resulting in a single field
> with the same aspect ratio as the frame. This may cause sampling
> artifacts, but I'm not sure.
You can subsample fields in the x direction, if your algorithms do not
need that amount of data. Using a simple [1 2 1] kernel does not yield
much aliasing effects on "standard" TV images ([1 2 1] is a triangle in
image space, thus a square sinc in Fourier).
--Nicolas
______________________________________________________________________
/\ \
\_| Nicolas Devillard | E-mail: nDevil@eso.org |
| European Southern Observatory, | http://www.eso.org/~ndevilla |
| Karl-Schwarzschild-Str. 2, | Phone: +49 89 320-06-340 |
| D-85748 Garching, Germany | Fax : +49 89 320-06-358 |
| __________________________________________________________________|_
\_/____________________________________________________________________/
Subject: Need help with min/max of fn. from Fourier Transform
From: Ryan Sparkes
Date: Thu, 14 Nov 1996 16:07:28 GMT
Hi. I need to know if there's a way to do the following:
I have some data representing a function F(u,v,w) that is the
Fourier transform of some unknown function f(x,y,z) (this second
function is a stack of images, and it is known to be composed
of entirely real data). Is there any way I can find the maximum
and minimum values of f(x,y,z) without doing the actual inverse
Fourier Transform? It would save me a lot of computing time
if there was a way.
Thanks,
RTS
ryans@isgtec.com
Subject: Sorry to ask for Compression in this NG.
From: Alexander Johannesen
Date: Thu, 14 Nov 1996 20:16:34 +0100
An urgent hi,
This is frustrating. I just finished my own compression
routines, but they are slow AND not very compressing :(
and within three days I have got to be finished with
this thing!
Anybody out there got a VERY SIMPLE compression routine
with source (ASM or C for MS-DOS) ? (Might even be that
LZH / GIF compression is more than good enough.)
What I got is ;
*ptr pointer to 64K buffer
*res pointer to 64K result buffer
all I need is ;
Compress the stuff in *ptr, and put
it compressed into *res, returning
the sise of compressed data.
This is for grayscale 8-bit images, with LOTS of room
for compression.
HELP! Can anybody help me?
Kind regards,
Alexander Johannesen (alejohan@sn.no) Norway
Subject: Postdoctoral Positions at NIST
From: Martin Herman
Date: Thu, 14 Nov 1996 16:04:12 -0500
The Perception Systems Group at the National Institute of Standards
and Technology (NIST) has postdoctoral openings in computer vision and
robotics. U.S. CITIZENSHIP IS REQUIRED for these positions.
NIST is a federal laboratory performing research in computer science,
manufacturing engineering (including robotics), electrical
engineering, applied math, chemistry, physics, materials science,
building and fire research, etc. NIST's mission lies in the
development of technology, measurements, and standards.
Currently, there are two robotics research programs in the Perception
Systems Group: autonomous outdoor vehicles and inspection of discrete
manufactured parts. The issues being investigated in the area of
autonomous vehicles include (1) integration of range images, camera
intensity images, and inertial positioning data for navigation,
obstacle avoidance, path planning, and obstacle and landmark
recognition, (2) the use of foveal-peripheral vision, active vision,
and gaze control, and (3) the integration of motion vision with range
data. Real-time performance on our outdoor vehicle testbed is a
critical issue.
In the area of manufacturing inspection, we are investigating the
integration of vision and active probing for inspecting the dimensions
of discrete manufactured parts (using both touch and non-contact
probes) and model-based recognition and registration of
parts. Real-time performance on our robotic inspection testbed is a
critical issue.
The postdoctoral appointment is for two years. The annual salary is
$45,500 plus full government benefits, relocation expenses, and $5,500
for professional travel, registration fees, books, etc.
The deadline for submitting a formal application with a brief research
proposal is January 15, 1997. To repeat, only U.S. CITIZENS may apply.
For further information, please contact:
Dr. Martin Herman
NIST
Building 220, Room B124
Gaithersburg, MD 20899
Tel: (301) 975-3441
E-mail: herman@nist.gov
http://isd.cme.nist.gov/staff/herman
Subject: Realtime Image analysis software for low end video camera
From: "N. Dhillon"
Date: Thu, 14 Nov 1996 16:29:03 -0600
Hello all,
I am looking for a software "platform" that will allow me to do realtime
image processing from a relatively low-end video input. For example, I
would like to find a camera (like Connectix Quickcam) with a
corresponding API to some Image Processing platform (perhaps Matlab
like) that would allow me to write a program to perform facial
recognition tasks (like detect when an eye is open, or closed, i.e. not
high resolution or high cost of error tasks). Ideally the package would
be PC (windows95) based, and would not be very expensive. Rate of
development is more important than speed and resolution.
Does anyone know of or have experience w/ such a platform/camera
combination or where to look?
Thank you,
Deep
Subject: NEEDED: B&W; Image Processing Toolkit/Code
From: Bryce
Date: Thu, 14 Nov 1996 14:31:57 -0800
I'm looking for source code (or MacOS PowerPC libraries) that supports
fairly simple image processing operations. In particular, I need to do
fast deskewing and despeckling on a 200dpi 1-bit/pixel scanned image.
Some higher-level image processing, such as feature analysis, might be
useful. And no, it has nothing to do with sending or receiving faxes,
though fax-related code would seem to be my best bet. I've found plenty
of libraries for Unix & Windows, but not much (quality) source code, and
almost no libraries. Any advice and/or suggestions would be greatly
appreciated.
Please direct all replies to:
Ken Krugler
ken@transpac.com
Subject: Re: image analysis system
From: Kia Peyvan
Date: Thu, 14 Nov 1996 15:30:08 -0800
matthew g donovan wrote:
>
> Our lab is considering purchasing an image analysis system for the
> quantification of immunohistochemical staining. Does anyone know of a
> good source for comparing the features of the currently available
> systems? We are looking for something compatible with our Nikon
> microscopes and very user freindly. We would welcome any suggestions of
> what to look for, as well as what to avoid. Thanks in advance, MD
Hi matthew, we are a company that makes imaging system for life
sciences and we are also developing imaging sytems for microscopy. Our
systems are based on win95 winNT operating sytems and are designed to be
very user friendly. We can also develope custom imaging application for
your specific needs.
Please feel free to contact us for further details,
Sincerely
-----------------------------------------------------------------------
Kia Peyvan
Software Development Manager
Life Sciences Division
NucleoTech Corporation 957 Industrial Rd, Suite R
San Carlos, CA 94070
Tel: 415-596-8747
Fax: 415-596-8814
nucleotech@earthlink.net
Subject: CfP: EuroPar'97 Workshop 10
From: Hartmut Schmeck
Date: Fri, 15 Nov 1996 08:22:33 +0100
========================================================================
=
Euro-Par'97
===========
Announcing a Euro-Par'97
Workshop
------------------------------------------------------------------------
-
Image and Signal Processing and Special-Purpose Processors
------------------------------------------------------------------------
-
Program Committee:
------------------
* Patrice Quinton, IRISA--CNRS, Rennes, France, Chair
* Hartmut Schmeck, University of Karlsruhe, Germany, Local Chair
* Ed F. Deprettere, Delft University of Technology, The Netherlands,
Vice-Chair
* Edward A. Lee, UC Berkeley, USA, Vice-Chair
Deadline:
---------
20 January 1997 (paper) 1 February 1997 (electronic)
General Description:
--------------------
Euro-Par is the annual European conference in Parallel Processing. Like
the 1996 conference in Lyon, the 1997 version will consist of a number
of highly focussed workshops on all aspects of parallel processing,
from theory to practice, from academy to industry. The workshops will
present the latest advances in their respective domains. In addition,
there will be a number of high-level tutorials of general interest plus
a series of invited talks. Calls for papers for 20 workshops are being
launched. All accepted papers will appear in the proceedings published
by Springer-Verlag in the LNCS Series.
Workshop Description:
---------------------
Workshop #10: Image and Signal Processing and Special-Purpose Processors
Our evolving information society increasingly relies on
efficient algorithms and architectures for Signal and Image Processing
(SIP). Due to its extremely high performance requirements, SIP has
always been one of the major application areas and one of the driving
forces behind the design of special-purpose processors. New challenges
arise from areas like computer vision or multi-media applications.
Submissions are invited from academia and industry on various aspects
of SIP: parallel architectures, methods for designing parallel
algorithms, languages and environments for programming parallel
architectures or for designing special-purpose architectures for SIP
applications, applications of parallel processing in SIP, interactions
between algorithms, architectures and their development and design
methodologies.
Topics of interest include:
* parallel architectures and algorithms
* dynamically reconfigurable architectures
* parallel DSP systems
* application specific parallel architectures
* parallel custom computing machines for SIP
* methods and tools for the design of SIP architecture
* SIP design environments
* computer vision
* parallel video servers
* signal and image compression
* dependability and security of SIP
Further information
-------------------
is available at URL http://www.uni-passau.de/europar97/. It includes
the list of all workshops. Please send all information requests and
comments to europar97@fmi.uni-passau.de. Register today on the
Euro-Par'97
mailing list by sending us a mail! See below for additional information.
Official Address:
-----------------
Euro-Par'97, Universitaet Passau, D-94030 Passau, Germany
Phone: (+49) (851) 509-3071; Fax: (+49) (851) 509-3092
E-mail: europar97@fmi.uni-passau.de
URL: http://www.uni-passau.de/europar97/
------------------------------------------------------------------------
-
ADDITIONAL INFORMATION
----------------------
Submissions are expected to be no longer than 12 pages in the LNCS style
(page size 12.2 cm x 19.3 cm, 12 pt Computer Modern font). Style files
are on the Web at http://www.springer.de/author/tex/help-tex.html (for
TeX click on PLNCS, for LaTeX on LLNCS). Text processors other than TeX
or LaTeX may be used, but we strongly suggest that the submissions
satisfy the LNCS specifications!
Submitted papers must not be simultaneously under review for any other
conference, and authors must point out any substantial overlap with
their previously published work.
Papers will be published in three categories: distinguished papers
(12 pages), full papers (8 pages), and short papers (4 pages). Authors
should state with their submission whether classification as a short
paper would be acceptable. Except in extenuating circumstances,
submission should be electronic.
! Electronic submissions are mandatory for all people having !
! access to standard electronic facilities. !
ELECTRONIC SUBMISSIONS should consist of two e-mails:
----------------------
* the first one containing an ASCII cover page, giving the author's
full name, address, phone and fax number, e-mail address,
a 100-word abstract and keywords, as well as the workshop number
(#10) and the workshop title (Image and Signal Processing and
Special-Purpose Processors),
* the second one with the printable file.
Accepted formats are uuencode'd compressed EPSF PostScript files (most
preferably generated by dvips).
! Important! We are sorry that we are unable to accommodate !
! other file formats. If at all possible, please attempt to !
! preview the document with Ghostview and print it on more !
! than one brand/model of laser printers. If your program !
! requests it, assume that the document will be printed on a !
! 300 DPI standard PostScript Unix Sun laser printer. We will !
! do our best to get all files printed. If a file cannot be !
! printed in spite of all our efforts, we will have to reject !
! the electronic submission. In case of doubt, send us a !
! (possibly preliminary) paper version of the submission !
! before the 20 January 1997 paper deadline (one copy is !
! enough in this case). We will assume that the electronic !
! version of the submission, if printable, is the final one. !
The files should be mailed to europar97@fmi.uni-passau.de and received
at this address before 1 February 1997. PLEASE NOTE THAT THIS IS A FIRM
DEADLINE.
PAPER SUBMISSIONS are only allowed for people without easy access to
standard electronic facilities. The cover page must clearly indicate
the name, address, phone and fax numbers and e-mail address of the
author(s), and contain a 100-word abstract and keywords. The workshop
number (#10) and the workshop title (Image and Signal Processing and
Special-Purpose Processors) relating to each submitted paper must also
be clearly indicated on the cover page. Authors should send six copies
of their submission to the official address, postmarked before 20
January
1997. PLEASE NOTE THAT THIS IS A FIRM DEADLINE.
Proceedings:
------------
All accepted papers will be published by Springer-Verlag in the
Lecture Notes in Computer Science Series. Proceedings will be
available at the conference.
Important Dates:
----------------
Register on the mailing list: Today!
Deadline for paper submissions: 20 January 1997
Deadline for electronic submissions: 1 February 1997
Notification of acceptance: 2 May 1997
Final version due: 10 June 1997
Early registration: 30 June 1997
Euro-Par'97 in Passau: 26--29 August 1997
------------------------------------------------------------------
Prof. Dr. Hartmut Schmeck E-Mail: schmeck@aifb.uni-karlsruhe.de
Institut AIFB Tel.: +49-721-608-4242
Universitaet Karlsruhe Fax: +49-721-693717
D-76128 Karlsruhe Besuchsanschrift: Englerstr.11 (R.219)
http://www.aifb.uni-karlsruhe.de/Staff/schmeck.html
------------------------------------------------------------------
Subject: Re: Optical Music Recognition
From: "JG.Campbell"
Date: Fri, 15 Nov 1996 15:31:53 +0000
Edoardo Volpi wrote:
>
> Dear Sirs,
>
> We are working on a important project to digitalize a large classic music's
> archive, and we are estimating the capability to utilize an Optical Music
> Recognition System in order to speed up the process. Such system must to be
> able to recognize also dynamical symbols.
[...]
> We would like to know your point of view on the matter and to check the
> potential cooperation on the project.
[...]
The following may give some idea of the research state-of-the-art:
O. Yadid-Pecht, M. Gerner, L. Dvir, E. Brutman, and U. Shimony,
Recognition of handwritten mucical notes by a modified Neocognitron,
Machine Vision and Applications (Springer-Verlag), No. 9, pp. 65--72,
1996.
Best regards,
Jon Campbell
--
Jonathan G. Campbell, ISC/ISE, University of Ulster, Magee College,
Derry, BT48 7JL, Northern Ireland. tel +44 1504 375367, fax 370040.
JG.Campbell@ulst.ac.uk http://www.iscm.ulst.ac.uk/~jon/
Subject: Re: Need help with min/max of fn. from Fourier Transform
From: lakshman@nsslsun.nssl.uoknor.edu (Valliappa Lakshmanan)
Date: 15 Nov 1996 15:23:58 GMT
In article <328B43C0.45BF@isgtec.com>, Ryan Sparkes wrote:
>Hi. I need to know if there's a way to do the following:
>
>I have some data representing a function F(u,v,w) that is the
>Fourier transform of some unknown function f(x,y,z) (this second
>function is a stack of images, and it is known to be composed
>of entirely real data). Is there any way I can find the maximum
>and minimum values of f(x,y,z) without doing the actual inverse
>Fourier Transform? It would save me a lot of computing time
>if there was a way.
>
>Thanks,
>
>RTS
>ryans@isgtec.com
Min and max are non-linear operators. I doubt that you can get at them
in a closed form. However, you _can_ get the lower and upper bound of
f(x,y,z).
Use the Schwarz inequality on the Fourier inverse transform equation:
sum(F(u,v,w).exp(...)) <= sum(abs(F(u,v,w) ) ) * total_num_points_in_uvw
not exactly the min,max but it might be enough for your application.
lakshman
