Copy Link
Add to Bookmark
Report
VISION-LIST Digest Volume 10 Issue 46
VISION-LIST Digest Fri Nov 01 12:31:58 PDT 91 Volume 10 : Issue 46
- 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:
Comparing Polygons
Info requested on corner detection
Characteristics of FLIR imagery
Karhunen-Loeve transform
Fast Hadamard Transform
Real time blob analysis boards?
Edge detection problem
Peano curves a.k.a. Millar polyhedron
Info. requested on Vehicle detection using vision....
Looking for CSE library
Neural Computing and Visual Industrial Inspection
----------------------------------------------------------------------
Date: Thu, 31 Oct 91 15:39:16 PST
From: csr!ibarroda@sol.UVic.CA (Ian Barrodale)
Subject: Comparing Polygons
In order to match corresponding points from two sets we have
(a) constructed all possible triangles in each set, and
(b) compared each triangle from the first set with every
triangle from the second set.
In making these comparisons, a geometrically invariant property called
sphericity (Pattern Recognition, 23(12), pp. 1333 - 1341, 1990) was
used. We now wish to use N-sided polygons rather than just triangles.
For, say, N=4 or 5, is there a similar, geometrically invariant
property that could be used to compare two polygons?
Ian Barrodale
ibarroda@csr.uvic.ca
------------------------------
Date: Sun, 27 Oct 91 21:32:49 +0100
From: yin@itk.unit.no
Subject: Info requested on corner detection
I am a Ph.D student. In one project, we need to detect corners as
some feature. I have tried some algorithm for corner detection.
But none is successful.
So, I would appreciate any suggestion, any reference book (papers),
and any personal research experince.I will summarize if I get sufficient
response.
Thanks,
Regards, Yin Shi
Division of Engineering Cybernetics : Phone +47 7 594376
The Norwegian Institute of Technology : Fax +47 7 594399
N-7034 Trondheim : Email yin@itk.unit.no
NORWAY
[ The Moravec detector is the classic. Others you've found useful?
phil... ]
------------------------------
Date: Mon, 28 Oct 91 19:28:48 PST
From: Gregory Heil <gheil@BBN.COM>
Subject: Characteristics of FLIR imagery
I have seen very little here on IR imagery over the past few
years. Last I saw, Phil was looking for test images ...
Can anyone point me towards answering the following questions:
1) What are IR CCDs in the various bands actually measuring?
2) How much (order of magnitude) control over "gain" and
"range" exists in military systems.
3) Where can I find tables of "target" emissivities
4) Test images ...
thanks .. GH
------------------------------
Date: Tue, 29 Oct 91 19:29:21 GMT
From: mulberry%triton.unm.edu@lynx.unm.edu ()
Organization: University of New Mexico, Albuquerque
Subject: karhunen-loeve transform
I am interested in obtaining any information on current software
and/or algorithms that deal with the karhunen-loeve transform. Has
there been any development in the way of "fast" algorithms?
Thanks.
mulberry@triton.unm.edu
------------------------------
Date: Thu, 31 Oct 91 17:19:12 -0500
From: slehar@park.bu.edu
Subject: Fast Hadamard Transform
Does anybody know of the existance of a fast hadamard transform,
preferably in the form of source code? If not, would you know about a
SLOW hadamard transform?
The Hadamard transform is like the Fourier transform except that its
basis function is a SQUARE wave, rather than a SINE wave.
------------------------------
Date: Thu, 31 Oct 91 08:34:39 +0100
From: smagt@fwi.uva.nl (Patrick van der Smagt)
Subject: real time blob analysis boards?
For my self-learning hand-eye coordination research, I am looking for
a piece of hardware that can do blob analysis in real time.
Specifically, the camera which is mounted in the manipulator of the
robot will see a black background with a few (often just one) white
objects in it. The object can be very small, just a few pixels, but
also very large, occupying 90% of the image. From each of the white
object in the image I need a few parameters: the x, y position, the
area, the width and height, and some such.
Are there any VME boards available that can do this in real time?
Patrick van der Smagt
University of Amsterdam
Dept. CS & M
Amsterdam, Netherlands
smagt@fwi.uva.nl
------------------------------
Date: 24 Oct 91 19:55:33 GMT
From: waynet@kit.cna.tek.com (Wayne Turner)
Organization: Tektronix, Inc., Redmond, Oregon
Subject: Edge detection problem
I previously posted this query to Vision-List@ads.com but it did not
appear so I'll try posting it directly to this news group.
[ The USENET network has had some flakiness. Please persevere if your
posting doesn't appear. phil... ]
Note that my query deals with a signal (not image) processing problem
but I am posting here since I think the problem may be amenable to
methods used in image processing. (Also I have not gotten much of a
response from previous posts to comp.dsp).
I want to develop an algorithm for labelling features in a
1-dimensional signal with equally spaced samples. The signal contains
about 2000-16000 points and the waveform is composed of linear segments
with negative slope. The segments are seperated by "faults" which are
5-50 samples long (but fixed length for a given signal). After each
fault, a new segment begins with roughly the same slope as the previous
segment but with a drop in amplitude. In other words, "featureless"
segments of the waveform are simply lines with negative slope (with
noise added).
The noise is non-Gaussian, has low and high-frequency components and
the noise amplitude increases as the signal amplitude decreases.
Attempts to model the noise well enough to make traditional signal
processing techniques work have failed.
Can anyone point me to references or have ideas regarding this
application: specifically, techniques for feature and edge labelling
(and threshold determination) in 1-dim. signals with non-stationary
noise.
Thank you,
Wayne Turner
waynet@kit.CNA.TEK.COM
------------------------------
Date: Wed, 30 Oct 1991 11:45:07 GMT
From: Tim Morris <dtm@ap.co.umist.ac.uk>
Subject: Peano curves a.k.a. Millar polyhedron
Does anyone have any clues (algorithms, references, etc) on how
to generate a three dimensional Peano curve (I've also seen it
called the Millar polyhedron)?
Thanks in advance,
Tim Morris (dtm@ap.co.umist.ac.uk)
------------------------------
Date: Thu, 31 Oct 91 20:14:20 -0600
From: Kishore Khemani <kishore@ee.umr.edu>
Subject: Info. requested on Vehicle detection using vision....
I am currently working on a project to simulate real time vehicle detection
and identification using vision.
I have set up a model with an RCA camera continously scanning the scene at
a traffic signal and am working towards developing an efficient and
reliable real time frame grabbing system with subsequent scene analysis
I would appreciate any info on research and working models and on any
papers on the subject.
Thanks
Kishore Khemani, UMR
------------------------------
Date: Thu, 24 Oct 1991 22:48:33 GMT
From: xinshen@acsu.buffalo.edu (Xin Shen)
Subject: Looking for CSE library
We need to use standard data of ECG. I wonder some people may tell me
where I can find it or if there are other sourses which can fulfil our
requirements.
Thank you.
------------------------------
Date: Thu, 24 Oct 91 10:43:46 GMT
From: eba@computing-maths.cardiff.ac.uk (Eduardo Bayro)
Organization: Univ. of Wales Coll. of Cardiff, Dept. of Electronic & Systems Engineering
Subject: Neural Computing and Visual Industrial Inspection
Dear Netters!
Recently many neuro-chips have been designed and built. Few authors
have presented neural nets applications for idustrial inspection.
Alexander developed WISARD using RAM based discriminators (neuro-RAM).
Two authors presented the applications of the back-propagation multi
layer perceptron and another the use of the Reduced Coulomb Energy
neural net for classification purposes.
These authors used accelerator boards ( ANZA or Balboa) to run fast
simulated nets. Can anybody give me his point of view on ways of
using neural computing for industrial visual inspection, considering cost
and current trends?. Are there any neural computers working on the
floor of a factory for classification and rejection of faulty objects?.
Is this a cost effective machine vision system?
Any references would be highly appreciated. Results will be posted back
to the list. Thanking you in advance, Eduardo.
------------------------------
End of VISION-LIST digest 10.46
************************
