🚀 go-pugleaf

Well,

Some of you may remember my early posting, regarding the intersection of
computer vision and logic programming.  For those of you who replied, many 
thanks.

What was found:

Alan J. Vayda (vayda@ee.ecn.purdue.edu) is using prolog for high-level
object recognition on range data. Ref: Kak, Vayda, Cromwell, Kim and Chen,
"Knowledge-Based Robotics", Proc. of the 1987 Conf. on Robotics and Auto-
mation.

Ray Reiter and Alan Mackworth at Univ. of British Colubia have a paper
"The Logic of Depiction" (UBC TR 87-24) which proposes a theory of vision
based in first order logic. Net address: mack%grads.cs.ubc.ca@RELAY.CS.NET.
(note: Dr. Mackworth, I could never get mail back to you. My address is at
the end of this message.)

In vol. 1 of "Concurrent Prolog: collect papers", edited by E. Shapiro, MIT
Press, 1987, S. Edelman and E. Shapiro have "Image Processing in Concurrent
Prolog", this deals with algoritms for low-level vision. (edelman or udi,
@wisdom.BITNET)

Denis Gagne's thesis, from U. of Alberta, in Edmonton, home of the Oilers
and the West Ed. Mall, describes a reasoning approach to scene analysis.
The person that refered this to me didn't know the title, but did give me
a lead on how to find out. Randy Goebel (goebel@alberta.uucp) and David
Poole (dlpoole@waterloo.csnet) were Denis's advisors.

Mulgaonkar, Shapiro and Haralick, describe a rule-based approach for
determining shap-from-perspective in "Shape from Perspective: a Rule Based
Approach", CVG&IP 36, pp. 289-320, 1986.

That's about all that I've heard.  I'm still interested in hearing more,
though.

Vince Kraemer (nosmo@cornu.ucsb.edu)
4946 La Ramada Dr.
Santa Barbara, CA 93111

"Always look on the bright side of life!" - man on the cross, in
                                                The Life of Brian
(note : our postnews has been ill, hence the delay.)