Summary: Image Segmentation and Labeling Using the Polya Urn Model
A. Banerjeey P. Burlinay F. Alajajiz
yDepartment of Electrical Engineering
University of Maryland at College Park
College Park, MD 20742­3275
and
zDepartment of Mathematics and Engineering
Queen's University
Kingston, ON K7L 3N6, Canada
Abstract
We propose a segmentation method based
on Polya's model for the spread of con­
tagion. An initial segmentation is ob­
tained using a Maximum Likelihood (ML)
estimate or the Nearest Mean Classifier
(NMC). The resulting clusters are then
subjected to a morphological process oper­
ating like the development of an infection
to yield segmentation of the image into ho­
mogeneous regions. This process is imple­

Source: Alajaji, Fady - Department of Mathematics and Statistics, Queen's University (Kingston)

Collections: Engineering