Copy Link
Add to Bookmark
Report
VISION-LIST Digest Volume 10 Issue 34
VISION-LIST Digest Fri Aug 09 09:37:49 PDT 91 Volume 10 : Issue 34
- 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:
Re: Deformable Objects
Retinex theory of color constancy
Public domain aerial and scenery images
Video cameras of high dynamic range
Tutorial Information on Contour Measurement
Image processing boards for Sparcs
Implementing Tsai's calibration
Looking for CCD camera
Need a Image Processing Package
Help with Autoview Viking Vision System
Research Associate - Vision by Associative Reasoning
Environmental Image Analysis Post-Doc Opportunity.
Conference Announcement: 7SCIA
Motion Workshop Agenda
----------------------------------------------------------------------
Date: Thu, 01 Aug 91 18:33:18 GMT
From: nagle@netcom.com (John Nagle)
Organization: Netcom - Online Communication Services UNIX System {408 241-9760 guest}
Subject: Re: Deformable Objects
Pentland's Thingworld is a fascinating system. Not only does
it have his deformation model, but it has finite-element analysis in
near-real-time, using a mathematically ingenious approach by which
the finite-element problem is translated into the frequency domain
and high-frequency terms dropped for speed.
Spend a few hours using Thingworld and you will have a much
clearer idea of what really happens when you push on something.
Nothing is really rigid; when you push on something it sends a
pressure wave through the object at its speed of sound. In Thingworld
you can watch this happen.
Thingworld points the way to the next generation of solid
modellers. It's a research tool, though; it doesn't support rigid or
stiff objects very well. If you try to create anything much stiffer
than Jell-o, the math tends to become unstable unless the step size is
made unusably small. There are some linearizing assumptions in the
math that are valid only for soft materials. The next step probably
requires a PhD-thesis sized improvement in the theory. So if anyone
is looking for a thesis topic and is into number-crunching and
graphics, I recommend this one.
John Nagle
------------------------------
From: milanese@cui.unige.ch
Date: 29 Jul 91 14:11
Subject: Retinex theory of color constancy
Hello,
I am working for a thesis in computer vision, trying to implement
Land and McCann's retinex theory of color constancy. I would be very
interested in knowing if there are already some public-domain
implementations of the theory, and to get in contact with other people
doing work on the same subject. Thanks a lot,
Thierry Vessereau
Computer Vision Group
Computer Science Center
University of Geneva
12, rue du Lac Fax: +41.22.7353905
CH - 1207 Geneva Telex: 423801 UNI CH
Switzerland E-mail: vesserea@cui.unige.ch
------------------------------
Date: 29 Jul 91 19:52:51 GMT
From: roy@osceola.cs.ucf.edu (Ian Roy)
Organization: University of Central Florida
Subject: HELP - public domain aerial and scenery images
Dear Netters,
I am looking for aerial and natural scenery images (preferably
in raw format) that is available in the public domain. I would
appreciate it if anyone could give me pointers to their locations.
Thanks,
Ian Roy
(roy@eola.cs.ucf.edu)
or
(air@pd2.ccd.harris.com)
------------------------------
Date: Tue, 30 Jul 91 19:38:31 GMT
From: rajan@rocdec.roc.wayne.edu (VARADARAJAN KRISHNAN)
Organization: Wayne State University, Detroit
Subject: video cameras of high dynamic range
Expires: When Boss fires me:-)
Keywords: video cameras, high contrast image capture
Hi,
I have a problem. We are trying to use image processing in the area of
medical physics application. For this we need to capture images
obtained from linear acceleratoars. The information in this image is
in terms of what is known as optical densities. Hihger the optical
density of the image lower is the percentage of light transmitted and
vice versa. To be able to interpret light intensity of pixels of
digitized image we use a calibration strip of steps of various optical
density values along with every image. The optical density strip
passes light ranging in percentages from 0.1% to 90% with 0.1% of
transmitted light corresponding to an optical density of 3.05 and 90%
of transmitted light corresponding to optical density 0.05.
(*&^%&*(*what is your point^&%$#*%#$...........!!!!:-) Now the
question The camera is unable to distinguish all the optical density
steps that is on the strip. I'm told i need a camera of high dynamic
range. How high ? Ours is a small establishment which is looking at
price ranges of not exceeding $10K. Anyone out there who can suggest
some pointers to handle such images???? Thanks much in advance.
Please directly respond to:
rajan@rocdec.roc.wayne.edu
varada@pacvax.roc.wayne.edu
------------------------------
Date: 1 Aug 91 04:17:31 GMT
From: huiliu@eecs.cs.pdx.edu (Hui Liu)
Subject: Tutorial Information on Contour Measurement
Organization: Portland State University, Portland, OR
Hi, Everybody:
This is Hui Liu in Portland State University. We are now doing
research on measuring the area of spots on human shin. The change
of area of spots is very important in detecting skin cancer. The
problems we are facing are:
Due to the curvature of human body, there is difference between
*----------------------*
the area we measured from the| two-dimension picture| we taken and
*----------------------*
the real area of the spots.
What we want to do now is try to find a efficient way to compensate
this error. It seems that we should measure the contour of human
body first. Then
1. How do we measure the contour? What's the most mature way to do
that?
2. What kind of useful information can we get from the contour
measurement?
3. How can we calculate the real area of spot based on this information?
Since we know very little about three-dimension image processing. Could
you please provide us some tutorial information, any book, any paper
on this issue so that we can find out where to start. Thanks.
------------------------------
Date: Tue, 30 Jul 91 16:54:27 SST
From: Shankar Narayanaswamy <ISSSN%NUSVM.bitnet%lilac.Berkeley.EDU@ucbvax.Berkeley.EDU>
Subject: Image processing boards for Sparcs
Apologies if this is the wrong newsgroup for this message, but it
is the closest I could find....
We are working on various aspects of image processing, including
image recognition, and would like to purchase image-processing
boards for Sparc 1+, IPC and 2 workstations. Preferably, these
boards would come with libraries of standard software. I would
appreciate:
1. Pointers to companies/products (FAX numbers would be nice)
2. Names of magazines that would carry advertisements and product
reviews of such boards, and,
3. Any other help/advice on locating such products.
We use X-windows and prefer the MOTIF window manager. We expect to
expand our work to include video in a year or two.
Thanks.
Shankar Narayanaswamy
isssn@nusvm.nus.sg
isssn@nusvm.bitnet
Institute of Systems Science, National University of Singapore
Phone : 772-6143 (voice), 778-2571 (fax)
If A + B = C, prove that D = E.
------------------------------
Date: Fri, 2 Aug 1991 16:02:41 GMT
From: damian@virtual.rose.utoronto.ca (Damian R. Kanarek)
Organization: University of Toronto, Engineering Computing Facility
Subject: Implementing Tsai's calibration
Keywords: Tsai, camera calibration
Hi there all,
I'm having some problems implementing Tsai's algorithm as published in
IEEE j.of robotics and automation (august 1987), and maybe some of you
out there could give me a hand.
The algorithm itself is pretty simple - only 1.5 pages long. It is
the notation he uses for the various coordinate systems which is
sometimes confusing, and probably the reason for the incorrect results
I'm getting.
1] When computing the distorted image coordinates he writes
"..grab a frame into the computer frame memory. Detect the row and column
number of each calibration point i. Call it (Xfi, Yfi)."
Ok, BUT if Xfi is the row number, that it will be the y-coordinate on the
screen! But if that's so, why is it called Xfi? Soo, which one is it?
Is Xfi the screen x-coordinate, or the y-coordinate???????????????
2] What is dy' (found in equation 7b and 8b)???? It just seems to pop out
of no-where! I know what dx is, I know dx', and dy. But dy' hasn't been
mentioned!
Thanks, and if you have a copy of the algorithm floating around, and if
it wouldn't be against some law for you to share it, please send me a copy!
Thanks a lot.
Damian
P.s. When responding, please include a "To Damian" at the top. Thanks.
------------------------------
Date: Fri, 9 Aug 1991 11:42:02 +0800
From: Marino Menozzi <menozzi@iha.ethz.ch>
Subject: looking for CCD camera
We are a researcher team at the Institute for Hygiene and Applied
Physiology at the Swiss Federal Institute of Technology (ETH) in
Zurich.
We are looking for a special CCD camera which enables to record
single frames. Exposure time should be controlled externally.
Furthermore the camera should allow a proper assignement of the
signal to each single camera pixel. Here are some detailed
specifications:
1. General specifications
1.1 Colour: Black / White
1.2 Resolution: 200 x 200 pixels or more
1.3 Costs: max a few thousand dollars
2. Exposure
2.1 Exposure shall be controlled from an external signal
2.2 Range of exposure time: 50 to 200 ms
2.3 Precision of start and stop of exposure: better than 0.5 ms
3. Readout
3.1 Analog readout of pixels at a rate inbetween 50 - 100 KHz
3.2 Readout pixel-clock generated by the camera and externally
avilable or pixel-clock generated externally
In case of a digital camera or a camera - frame grabber system
following specifications should be met:
- 256 gray levels
- proper assignement of data to single camera pixels
- PC AT compatible system
We would greatly apreciate if somebody helps us to find a system
with the specifications mentioned abouve. Informations may be
sent to following addresses:
e-mail: menozzi@iha.ethz.ch
fax# CH - 1 - 262 41 78, att. Urs Meile or Marino Menozzi
------------------------------
Date: Thu, 25 Jul 91 19:23:22 GMT
From: cshafai@arda.EE.UManitoba.CA (Cyrus Shafai)
Organization: University of Manitoba, Winnipeg, Canada
Subject: Need a Image Procesing Package
I don't know if I should post this in this group, so if it is not
relavent to this group please tell me what group to post this in.
I am looking for an image processing package that will run on a
Sun Sparc station IPC (or Sparc 1+, or Sparc 2). I require image
processing abilities such as
Contrast and Brightness adjustment
Edge enhancement
Scaling
Smoothing
Other features that I would like the system to have, although not as
important as the above features are
Fast Fourier Transform (FFT) ability
Some filtering (wiener filters, butterworth filters, etc.)
Log displays
Color display
Convolution
Printing ability (on a sparc printer)
Simple paint tools it would be nice
Thanks in advance
Cyrus Shafai <cshafai@eeserv.ee.umanitoba.ca>
The Canadian Scanning Tunneling Microscopy Laboratory
------------------------------
Date: 6 Aug 91 16:10:31 GMT
From: "D.G. Simon" <dgs12@eng.cam.ac.uk>
Subject: Help with Autoview Viking Vision System
AUTOVIEW VIKING VISION SYSTEM
Version 3 running on a RX02 Digital (DEC) computer.
RT11 operating system on 8 inch floppy disks
Using a Colorado Video Inc. framestore
Unfortunately our systems disk and backup have been corrupted.
Does anybody have any suitable software / hardware /
documentation to help us out in particular an RX02 systems disk?
Please contact Gideon Simon at: dgs12@uk.ac.cam.phx
(Cambridge University Engineering Dept.
Trumpington St. Cambridge CB2 1PZ)
------------------------------
Date: 30 Jul 1991 14:45:42 GMT
From: erh@minster.york.ac.uk
Subject: Research Associate - Vision by Associative Reasoning
UNIVERSITY OF YORK
Department of Computer Science
Research Associate - Vision by Associative Reasoning
An opportunity exists for a research associate to join an
active group investigating the use of connectionist
approaches for scene interpretation. The successful appli-
cant will be involved in developing hierarchical and multis-
cale methods for processing uncertain image data. This
research forms part of a collaborative project involving
British Aerospace, Defence Research Agency (formerly RSRE) -
Malvern, and the University of Surrey. The overall goal is
to understand how to construct parallel distributed systems
for visual information processing and to apply them to
aerial scene interpretation.
Although the post would suite an individual with a back-
ground in any mathematical or physical science, experience
of computer vision, image processing, neural networks or
pattern recognition would be of particular value.
This post is funded under the DTI-SERC Information
Engineering Programme; The appointment will be for a period
of 21 months and will be on the RA1A scale.
Informal enquires about this post can be made to Dr Edwin
Hancock on 0904 43 3374. Applications consisting of two
copies of a full curriculum vitae and the names of two
referees should be sent to the Personnel Officer, University
of York, York Y01 5DD to arrive no later then 16 August
1991.
------------------------------
Date: Wed, 7 Aug 91 10:01:28 -0700
From: wpkege@snll-arpagw.llnl.gov (kegelmeyer william p)
Subject: Environmental Image Analysis Post-Doc Opportunity.
Sandia National Laboratory's effort in environmental modeling is
dedicated to the acquisition and exchange of fundamental knowledge
that will improve the nation's assessment of factors in global climate
change. We are seeking qualified applications (U.S. citizens only)
for participation in this effort through the Sandia Postdoctoral
Program.
The problem of current interest is the modeling of 3D cloud shapes
from optical data gathered concurrently and sequentially from two or
more fish-eye lens cameras, with the first goal being the passive
extraction of cloud-bottom height measurements from this data. This
work will be extended to the extraction of other cloud parameters, to
application to general plume structures (in collaboration with
Sandia's Combustion Research facility), and to analysis of single-lens
data gathered by a moving remotely piloted vehicle.
The successful candidate will either have a background in image
analysis and an interest in learning the necessary meteorology, or a
background in meteorology coupled with strong experience and training
in image analysis.
Required experience includes:
* PhD in CS/EE or Meteorology.
* Knowledge of and hands-on experience with image processing and analysis
methods, particularly optical flow.
* Working knowledge of UNIX and C programming.
* Familiarity with Sun workstations is a plus.
Sandia National laboratory provides excellent facilities (a
workstation at every desk), a stimulating work environment, and
competitive postdoctoral salaries. Sandia is an Equal Employment
Opportunity/Affirmative Action employer, and minority candidates are
strongly encouraged to respond.
Please send your resume by August 31, 1991, to:
wpkege@sandia.llnl.gov (ASCII or PostScript only), or,
Philip Kegelmeyer, ORG 8351
Sandia National Laboratory
P.O. Box 969
Livermore, CA, 94551-0969
------------------------------
Date: Tue, 6 Aug 1991 16:27:26 +0100
From: hic@vision.auc.dk (Henrik I Christensen)
Subject: Conference Announcement: 7SCIA
7Scandinavian Conference on Image Analysis, 13-16 August 1991, Aalborg
You are kindly reminded and cordially invited to this conference on Image
Processing, Pattern Recognition and Computer Vision.
The conference language is English, and it covers the following themes:
Image Processing and Analysis Pattern Recognition
Perception Parallel Algorithms
Processor Architectures Image Scene Interpretation
Computer Vision Motion Analysis
Remote Sensing Industrial, Biological and Medical Applications
Invited talks and speakers:
Local Image Structure
Prof. Jan J. Koenderink, Utrecht Universitet, Holland
Towards Animate Vision at High, Intermediate and Low Levels
Prof. Christopher Brown, University of Rochester, USA
An Example of Learning in Knowledge-Directed Vision
Prof. Allen Hanson, University of Massachussetts, USA
Towards Continuously Operating Integrated Vision Systems for Robotics
Prof. James L. Crowley, Institut National Polytechnique Grenoble,
Frankrig
Remote Sensing for Thematic Mapping
Dr. Rupert Haydn, Gesellschaft fur Angewandte Fernerkenning
GmbH, Munich,
Germany
Shape Recognition in Mind and Brain
Prof. Irving Biederman, University of Minnesota, USA
Learning and Lessons Learned" in Computer Vision
Prof. Ruzena Bajcsy, University of Pennsylvania, USA
The 38 sessions of the conference cover further 106 papers and 35 posters.
All contributions are published in the two-volume conference proceedings
(1200 pages). An exhibition with leading suppliers of the field forms part
of the conference.
Registration: 7SCIA Conference Secretariat, c/o Aalborg Tourist Bureau,
steraa 8, DK-9000 Aalborg, Denmark. Phone: +45 98 12 63 55, fax: +45 98
16 69 22. The conference fee of DKK 3,000 includes lunches, refresehments,
welcome party and conference banquet (the latter not included if registered
after 12 Aug. noon).
ONE-day participation is possible on 15 Aug., "The Application & Exhibition
Day", with application oriented sessions and Prof. Crowley's invited talk.
Fee for this particular day only is DKK 1,400 (1,900 DKK incl.
proceedings).
Information and Programme: Laboratory of Image Analysis, Institute of
Electronic Systems, Aalborg University, phone: +45 98 15 85 22 ext. 4940,
fax: +45 98 15 40 08, (The secretariat of Conference Chair Professor Erik
Granum).
Dr. Henrik I Christensen | E-mail: hic@vision.auc.dk
Laboratory of Image Analysis | Phone: +45 98 15 11 33
Institute of Electronic Systems | Telefax: +45 98 15 40 08
Aalborg University | Home: Herluf Trollesgade 1, 2.th.
Fr. Bajers Vej 7, Bldg D1 | DK-9000 Aalborg, Denmark
DK-9220 Aalborg East, Denmark | Phone: +45 98 11 15 42
------------------------------
Date: Tue, 6 Aug 91 10:19:14 EDT
From: anandan@grumpy.sarnoff.com (P. Anandan )
Subject: Motion Workshop Agenda
ADVANCE PROGRAM
IEEE WORKSHOP ON VISUAL MOTION
IEEE SPECIAL WORKSHOP ON PASSIVE RANGING
OCT 7-10, 1991
NASSAU INN, PRINCETON, NEW JERSEY
GENERAL CHAIR PROGRAM CO-CHAIR & INFO. PROGRAM CO-CHAIR
Thomas S. Huang Peter J. Burt Edward H. Adelson
University of Illinois David Sarnoff Research Ctr. MIT Media Lab
CN 5300, Princeton, NJ 08543
Ph: (609)-734-2451
Fax: (609)-734-2313
Email: burt@vision.sarnoff.com
SPECIAL SESSIONS CHAIR IMAGE DATABASE
P. Anandan Rajesh Hingorani
David Sarnoff Research Ctr. David Sarnoff Research Ctr.
CN 5300, Princeton, NJ 08543 CN 5300, Princeton, NJ 08543
Ph: (609)-734-3249 Ph: (609)-734-2967
Email: anandan@vision.sarnoff Email: rh@vision.sarnoff.com
LOCAL ARRANGEMENTS
Ajit Singh Alice Archer
Siemens Corporate Res. David Sarnoff Research Ctr.
Princeton, NJ 0854 CN 5300, Princeton, NJ 08543
Ph: (609)-734-3389 Ph: (609)-734-2184
Email: singh@demon.siemens.com Email: Alice_Archer@maca.sarnoff.com
TECHNICAL COMMITTEE
Narendra Ahuja, U. Illinois Berthold Horn, MIT
Rama Chellappa, U. Maryland Kenji Mase, NTT, Japan
Ernst Dickmanns, Edward Riseman, UMass
Univ. der Bundeswehr, Munchen Shmuel Peleg, Hebrew Univ.
David Fleet, Queens Univ., Ontario Brian Schunck, U. Michigan
Gosta Granlund, Linkoping Univ. Saburo Tsuji, Osaka Univ.
Keith Hanna, David Sarnoff Res. Ctr. Martin Vetterli, Columbia Univ.
David Heeger, NASA Ames Andrew Watson, NASA Ames
IEEE SPECIAL WORKSHOP ON PASSIVE RANGING
Sponsored by NASA and DARPA
OCT 10, 1991
Workshop Co-Chairs
Bir Bhanu Banavar Sridhar
Univ. of California NASA Ames Research Center
Riverside, CA 92521 Moffett Fields, CA 94035
Ph: (714)-787-3954 or (714)-787-5190 Ph: (415)-604-5450,
Email: bhanu@ucrmath.ucr.edu Email: bsridhar@ames.arc.nasa.gov
IEEE WORKSHOP ON VISUAL MOTION
PRELIMINARY AGENDA
MONDAY - OCTOBER 7, 1991
8: 30 AM WELCOMING REMARKS
8: 50 AM STRUCTURE & MOTION FROM EXTENDED SEQUENCES
Passive navigation in a partially known environment
S. Chandrashekhar and R. Chellappa, U. of Southern California, USA
Incorporating motion error in multi-frame structure from motion
J. Oliensis and J. Inigo Thomas, U. of Mass., USA
Structure and motion from uniform 3D acceleration
W. Franzen, U. of Southern California, USA
Factoring image sequences into shape and motion
C. Tomasi and T. Kanade, Carnegie-Mellon U., USA
10: 10 AM COFFEE BREAK
10: 40 AM ANALYSIS OF IMAGE FLOW - 1
Parallel visual motion analysis using multiscale Markov random fields
F. Heitz, P. Perez, E. Memin and P. Bouthemy, IRISA/INRIA, France
Incremental estimation of image-flow using a Kalman filter
A. Singh, Siemens Corporate Research, USA
Estimation of velocity, acceleration and disparity in time sequences
H. Barman, L. Haglund, H. Knutsson and G.H. Granlund, Linkoping U., Sweden
Stability of phase information
D.J. Fleet and A.P. Jepson, Queen\'d5s U., Canada
12: 00 Noon LUNCH
1: 40 PM COMBINED MOTION AND STEREO
Estimating three-dimensional vehicle motion in an outdoor scene from
monocular and stereo image sequences
M.K. Leung, Y. Liu and T.S. Huang, U. of Illinois at Urbana-Champaign, USA
Recovering a boundary-level structural description from dynamic stereo
A.P. Tirumalai, B.G. Schunck and R.C. Jain, U. of Michigan, USA
Motion and structure from long stereo image sequences
N. Cui, J. Weng and P. Cohen, Ecole Polytechnique de Montreal, Canada
Robustness of structure from binocular known motion
R. Dutta and M.A. Snyder, U. of Mass., USA
3: 00 PM COFFEE BREAK
3: 30 PM MODELS OF HUMAN & BIOLOGICAL VISION
Motion-boundary illusions and their regularization
Y. Aloimonos and L. Huang, U. of Maryland, USA
Model of cell responses in visual area MT
D.J. Heeger, A.D. Jepson and E.P. Simoncelli, NASA-Ames, USA
Dynamic reconstruction of 3D structure and 3D motion
H. Ando, MIT, USA
An adaptive multi-scale approach for estimating optical flow:
computational theory and physiological implementation
C. Koch, H.T. Wang, R. Battiti, B. Mathur and C. Ziombowski, California
Institute of Technology, USA
8: 00 PM EXPERIMENTAL RESULTS SESSIONS
A. Structure and Motion --- Chair: Prof. Rama Chellappa, USC
B. Flow --- Chairs: Dr. P. Anandan, Sarnoff and Dr. Ajit Singh, Siemens
C. Stereo and Motion --- Chairs: Prof. Narendra Ahuja, U. of Illinois
TUESDAY - OCTOBER 8, 1991
8: 30 AM RECOVERY OF EGO-MOTION
Approximate subspace methods for recovering rigid motion
A.D. Jepson and D.J. Heeger, U. of Toronto, Canada
Motion recovery from image sequences using only first order optical flow
information
S. Negahdaripour and S. Lee, U. of Hawaii, USA
Ego-motion from global flow field
V. Sundareswaran, New York U., USA
Direct computation of the focus of expansion from velocity field
measurements
R. Guissin and S. Ullman, Weizmann Institute, Israel
Direct multi-resolution estimation of ego-motion and structure from
motion
K.J. Hanna, David Sarnoff Research Center, USA
10: 10 AM COFFEE BREAK
10: 40 AM ANALYSIS OF MULTIPLE MOTION
Principle of superposition: a common computational framework for
analysis of multiple motion
M. Shizawa and K. Mase, NTT, Japan
Robust estimation of a multi-layered motion representation
T. Darrell and A. Pentland, MIT, USA
Factorization-based segmentation of motions
T.E. Boult and L.G. Brown, Columbia U., USA
Mechanisms for isolating component patterns in the sequential analysis
of multiple motion
P.J. Burt, R. Hingorani and R.J. Kolczynski, David Sarnoff Research Center
12: 00 Noon LUNCH
1: 40 PM ROBOTICS AND NAVIGATION
Relative depth from motion using normal flow: an active and purposive
solution
L. Huang and Y. Aloimonos, U. of Maryland, USA
Realizing active vision by a mobile robot
S. Li, I. Myawaki, H. Ishiguro and S. Tsuji, Osaka U., Japan
Quantitative interpretation of image velocities in real time
J. Huber and V. Graefe, Univ. der Bundeswehr Muenchen, Germany
A new approach to vision and control for road following
D. Raviv and M. Herman, National Inst. of Standards and Technology, USA
Direct estimation of time-to-impact from optical flow
M. Tistarelli and G. Sandini, U. of Genoa, Italy
3: 20 PM COFFEE BREAK
3: 50 PM STRUCTURE & MOTION FROM MULTIPLE VIEWS
Getting around the aperture problem using point and line correspondences
M. Spetsakis, York U., Canada
Structure and motion in two dimensions from multiple images, a least
squares approach
C.J. Taylor, D.J. Kriegman and P. Anandan, Yale U., USA
Towards structure from motion for linear features through reference
points
L. Quan and R. Mohr, LIFIA-IMAG, France
8: 00 PM EXPERIMENTAL RESULTS SESSIONS
A. Ego-Motion --- Chair: Dr. David Heeger, NASA Ames
B. Multiple Motion --- Chair: Prof. Shmuel Peleg, Hebrew Univ.
C. Robotics and Navigation --- Prof. Edward Riseman, UMass
WEDNESDAY - October 9, 1991
8: 30 AM OBJECT SHAPE AND SEGMENTATION FROM MOTION
Simultaneous estimation of 3D shape and motion of objects by computer
vision
J. Schick and E.D. Dickmanns, U. of Bundeswsehr Muenchen, Germany
Forming a three dimensional environment model using multiple
observations
P. Khalili and R. Jain, U. of Michigan, USA
3D structure reconstruction from an ego-motion sequence using
statistical estimation and detection theory
Y.-L. Chang and J.K. Aggarwal, U. of Texas, Austin, USA
Segmentation, matching and estimation of structure and motion of
textured piecewise planar surfaces
S. Sull and N. Ahuja, U. of Illinois at Urbana-Champaign, USA
Hierarchical feature-based matching for motion correspondence
V. Venkateswar and R. Chellappa, U. of Southern California, USA
10: 10 AM COFFEE BREAK
10: 40 AM NON-RIGID MOTION
Non-rigid motion and structure from contour
A. Pentland, B. Horowitz and S. Sclaroff, MIT, USA
Pointwise tracking of left-ventricular motion
A.A. Amini and J.S. Duncan, Yale U., USA
Motion analysis and modeling of epicardial surfaces from point and line
correspondences
S.K. Mishra and D.B. Goldgof, U. of Southern Florida, USA
Recursive estimation of non-rigid shape and motion
D. Metaxas and D. Terzopoulos, U. of Toronto, Canada
12: 00 Noon LUNCH
1: 00 PM ANALYSIS OF IMAGE FLOW - 2
comparison among four approaches
C. Koch, A. Moore, W. Bair, T. Horiuchi, B. Bishofberger and J. Lazzaro
California Institute of Technology, USA
Segmentation of people in motion
A. Shio and J. Sklansky, NTT, Japan
Adaptive coring techniques for spatio-temporal signals
J. Lubin, David Sarnoff Research Center, USA
Motion tracking on the spatiotemporal surface
H.H. Baker, SRI International, USA
2: 30 PM VISITS TO LABORATORIES IN THE PRINCETON AREA
**************************************************************************
IEEE SPECIAL WORKSHOP ON PASSIVE RANGING
OCTOBER 10, 1991
PRELIMINARY AGENDA
8: 15 AM Introduction NASA
8: 30 AM Recursive Range Estimation Using Sequence of Helicopter
Images
Banavar Sridhar, NASA Ames, Ca., USA
9: 05 AM Building 3-D Environmental Models by Reasoning by Doing
Ramesh Jain, Univ. of Michigan, USA
9: 40 AM Segmentation and Range Estimation by Feature-Based and
Region-Based Stereo
Tom Binford, Stanford Univ., Ca., USA
10: 15 AM BREAK
10: 30 AM The Application of 3-D Pose Determination Techniques to 3-D
Model Extension
Allen Hanson and Rakesh Kumar, Univ. of Massachusetts, USA
11: 05 AM 3-D Motion Estimation and Object Recognition from Stereo
Image Sequences
Tom Huang, Univ. of Illinois, USA
11: 40 AM Experimental Results and Multibaseline and Adaptive Window
Stereo
Takeo Kanade, Carnegie Mellon Univ., Pa., USA
12: 15 PM LUNCH
1: 30 PM INS Integrated Passive Ranging System for Obstacle
Detection
Bir Bhanu and Barry Roberts, UC Riverside and Honeywell
2: 05 PM Visual Ranging for Road Vehicle Navigation
Ernst Dickmanns, Univ. der Bundeswehr, Munchen, Germany
2: 40 PM The DROID 3-D Vision System for Temporal Integration of
Stereo and Motion
Chris Harris and Mike Stephens, Roke Manor Research, UK
3: 15 PM BREAK
3: 30 PM Viewpoints on Issues in Passive Ranging
(Each speaker presents his viewpoints for 15 minutes)
John (Yiannis) Aloimonos, Univ. of Maryland
Larry Matthies, JPL
Marty Herman, NIST
Rama Chellappa, USC
4: 30 PM PANEL DISCUSSION
5 panelists from the list of speakers
**********************************************************************
Workshop Location
The workshop will be held at Nassau Inn at Palmer Square in Princeton, NJ. A
block of rooms has been reserved for the workshop participants at the rate of
$85 single, and $125 double. Please contact Nassau Inn at (609)-921-7500 and
don't forget to identify yourself as an IEEE workshop participant. A hotel
registration form is enclosed at the end of this program.
Transportation
By Air:
Newark International Airport is the most convenient airport for Princeton.
The Princeton Airporter (telephone: (609)-587-6600) offers frequent limousine
service to the Nassau Inn from the Newark Airport . They also service JFK
Airport, New York via a connecting shuttle from JFK to Newark. The fare is
approximately $19 from Newark Airport and $29 from JFK Airport.
By Train:
Take the train (AMTRAK or New Jersey Transit Northeast Corridor Line) to
Princeton Junction and take the local shuttle to Princeton. The hotel is less
than a mile from the Princeton station. Alternatively stop at Trenton and take
a cab or rent a car.
By Car:
Coming from the north: Take New Jersey Turnpike Southbound to Exit 9 (New
Brunswick). Follow signs to Route 1 South (towards Trenton). Travel about 15
miles on Route 1 and make a right turn at the sign for Princeton Hightstown --
Princeton Business District (Route 571/526). Take this road, which becomes
Washington Road, for about 5 minutes to Nassau St. (Route 27). Make a left
onto Nassau St. Go three blocks and turn right (at White Gazebo) onto Palmer
Square and the Nassau Inn. (Public parking is available at the Chambers St.
Garage.)
Coming from the south: Take New Jersey Turnpike Northbound to Exit 8
(Hightstown). Bear to left to Route 33 West. Follow 33 West to Firehouse in
Hightstown; make a left turn on Main St. Make a right turn (at light) onto
Stockton St. Stay on Stockton. Cross Route 130 onto Princeton-Hightstown
Road which is Route 571 West. After Rte. 571 crosses Rte. 1, it becomes
Washington Road. Follow Washington Road to Nassau St. (Route 27). Make a
left onto Nassau St. Go three blocks and turn right (at White Gazebo) onto
Palmer Square and the Nassau Inn. (Public parking is available at the
Chambers St. Garage.)
Workshop Registration Information
The workshop registration fee covers attendance at all the sessions of the
workshop on visual motion, refreshments during breaks, and the workshop
proceedings, which will be available at the workshop. A separate
registration fee will be charged for the special workshop on Passive Ranging.
The deadline for pre-registration is Sept. 13, 1991. A registration form is
enclosed with this program.
Special Experimental Sessions
A novel feature of the motion workshop is the special experimental sessions.
These sessions will include presentations of experimental results on images
available in the workshop database of images, and will be generally geared
towards discussion of issues related to algorithm performance and other
practical issues. These sessions will take place between 8-10 PM on Oct 7 and
8. Six such sessions are currently planned. If you would like to participate
in one of these sessions please contact the organizer of these sessions, P.
Anandan or one of session chairs, whose names are listed in the detailed
program enclosed. Final selection of participants will be made by the
respective session chairs. For details on the image database, please contact
Rajesh Hingorani.
Other Hotels in the Area
Hyatt Regency, Princeton. Tel.: (609)-987-1234
Palmer Inn, Princeton. Tel. (609)-452-2500
Ramada Inn, Princeton. Tel. (609)-452-2400
Marriott Hotel, Princeton. Tel. (609)-452-7900
Please contact Alice Archer (Local Arrangements) for information concerning
student accommodations and other options.
**************************************************************
IEEE WORKSHOP ON VISUAL MOTION
HOTEL RESERVATION FORM
Return to: NASSAU INN, Palmer Square, Princeton, NJ 08542 (609) 921-7500
A block of rooms is being held for the workshop participants until three weeks
before the workshop at the special rates of $85 single, and $125 double.
Please reserve your room early. Check-in time is 6 PM. To guarantee your
room in case of late arrival, you must pay one night's lodging at the time of
making your reservation. While making telephone reservations by telephone be
sure to identify yourself as a IEEE workshop participant. Payment may be made
by check or by providing appropriate credit card information.
NAME:
ORGANIZATION:
STREET:
CITY/STATE/ZIP
TYPE OF ROOM (SINGLE/DOUBLE):
ARRIVAL DATE AND TIME: DEPARTURE DATE:
CREDIT CARD NAME:
CREDIT CARD #: EXPIRATION DATE:
CARDHOLDER'S NAME:
***********************************************************
IEEE WORKSHOP ON VISUAL MOTION
IEEE SPECIAL WORKSHOP ON PASSIVE RANGING
October 7-10, 1991
WORKSHOP REGISTRATION FORM
Return this form with a check, purchase order (U.S. dollars) or credit card
endorsement payable to:
IEEE Computer Society,
1730 Massachusetts Ave. N.W.,
Washington, DC 20036-1903
Ph (202) 371-1013, FAX (202) 728-0884
NAME:
ORGANIZATION:
STREET:
CITY/STATE/ZIP:
WORK PHONE NUMBER:
FAX NUMBER:
IEEE MEMBERSHIP NUMBER:
CREDIT CARD #: EXPIRATION DATE:
CARDHOLDER'S NAME:
CARDHOLDER'S SIGNATURE: DATE:
REGISTRATION FEES (Please circle appropriate fees):
MOTION WORKSHOP
Member Non-Member Student
Until 9/13/91 $190 $255 $75
After 9/13/91 $230 $295 $85
(late/onsite)
PASSIVE RANGING WORKSHOP
Member Non-Member Student
Until 9/13/91 $40 $50 $30
After 9/13/91 $50 $65 $35
(late/onsite)
TOTAL FEES ENCLOSED:
MODE OF PAYMENT: (Circle one)
American Express Visa Master Card
Personal Check Company Check Purchase Order
NOTE: Written requests for refunds must be received in the IEEE Computer
Society office no later than 9/13/91. Refunds are subject to a $50 processing
fee. All no-show registrations will be billed in full. Students are required
to show current picture ID at the time of registration. Registrations after
9/20/91 will be accepted on-site only.
------------------------------
End of VISION-LIST digest 10.34
************************
