Mean field decomposition of a posteriori probability for MRF-based unsupervised textured image segmentation.
Hideki NodaMehdi N. ShiraziBing ZhangEiji KawaguchiPublished in: ICASSP (1999)
Keyphrases
- markov random field
- image segmentation
- textured images
- graph cuts
- mrf model
- random fields
- belief propagation
- energy function
- maximum a posteriori
- parameter estimation
- loopy belief propagation
- higher order
- energy minimization
- unsupervised clustering
- unsupervised learning
- image decomposition
- cluster validation
- supervised learning
- input image
- supervised classification
- pairwise
- markov networks
- belief networks
- conditional random fields
- segmentation algorithm
- normalized cut
- semi supervised
- texture image segmentation
- shape prior
- region growing
- active contours
- level set
- probability distribution
- multiresolution
- multiscale
- image processing
- object segmentation
- decomposition method
- bayesian networks
- unsupervised image segmentation
- computer vision