DocumentCode
1335814
Title
MAP-based probabilistic diffusion method for correspondence and line field estimation
Author
Lee, Sang Hwa ; Jong-Il Park ; Lee, Choong Woong
Author_Institution
Image Commun. Lab., Seoul Nat. Univ., South Korea
Volume
36
Issue
7
fYear
2000
fDate
3/30/2000 12:00:00 AM
Firstpage
627
Lastpage
628
Abstract
A new maximum a posteriori (MAP)-based probabilistic diffusion method is proposed for the estimation of dense correspondence and line fields. The proposed algorithm reflects the joint probabilistic distributions of the neighbourhood in the Markov random field, and implements them by using a plane configuration model. The probabilistic diffusion scheme showed fast convergence and improved object boundaries in the estimated fields
Keywords
Markov processes; image processing; maximum likelihood estimation; MAP probabilistic diffusion method; Markov random field; convergence; correspondence estimation; image processing; line field estimation; maximum a posteriori algorithm; object boundary; plane configuration model;
fLanguage
English
Journal_Title
Electronics Letters
Publisher
iet
ISSN
0013-5194
Type
jour
DOI
10.1049/el:20000398
Filename
842203
Link To Document