🚀 go-pugleaf

       DEPARTMENT OF COMPUTER SCIENCE
          COLORADO STATE UNIVERSITY

                 BARNEY'S
        MONDAY AFTERNOON CLUB SEMINAR

     Speaker:	Dr. Bruce A. Maxwell
        	University of North Dakota

	Date:	Monday, March 17, 1997

	Time:	3:30 - 4:30 p.m.

	Place:	100 USC

	Refreshments:  following the talk

	Hosts:	Stephen B. Seidman, Ross Beveridge
		seidman@cs.colostate.edu, 
		ross@cs.colostate.edu

      Segmentation and Interpretation Using Multiple
	    Hypotheses of Image Formation

	One of the first and most important tasks in 
single image analysis is segmentation: finding groups 
of pixels in an image that "belong" together.  A seg-
mentation specifies regions of an image that we can 
reason about and analyze. An accurate segmentation is 
a prerequisite for vision tasks such as shape-from-
shading.  A general-purpose segmentation algorithm, 
however, does not currently exist.  Furthermore, the 
output of many segmentation algorithms is simply a 
set of pixel groupings; no attempt is made to provide 
a physical description of or make a connection between 
the image regions and objects in the scene.

	Physics-based segmentation algorithms are 
based upon identifying coherent regions of an image 
according to a model of object appearance.  These 
models have usually assumed that a scene contains a 
single material type, restricted forms of illumina-
tion, and uniformly colored objects. This talk will 
present a new framework that challenges these assump-
tions by considering multiple physical hypotheses 
for simple image regions. 

	A consequence of this framework is a new 
approach to segmenting complex scenes into regions 
corresponding to coherent surfaces rather than merely 
regions of similar color. Furthermore, by selecting 
among the hypotheses considered for each region, we 
obtain not only the spatial grouping, but also an 
abstract interpretation of each aggregate region. 
This talk will focus on an implementation of this 
new approach to segmentation and present segmenta-
tions and the corresponding interpretations of 
scenes containing multi-colored piece-wise uniform 
objects.