Data-driven segmentation of textured images using hierarchical markov random fields.
Hideki NodaMehdi N. ShiraziPublished in: Systems and Computers in Japan (1995)
Keyphrases
- markov random field
- segmentation of textured images
- data driven
- textured images
- multiresolution
- random fields
- texture segmentation
- higher order
- maximum a posteriori
- belief propagation
- image segmentation
- parameter estimation
- image restoration
- mrf model
- graph cuts
- energy minimization
- pairwise
- energy function
- conditional random fields
- potts model
- natural images
- potential functions
- low level vision
- partition function
- map estimation
- loopy belief propagation
- markov networks
- texture features
- color images
- discriminative random fields
- approximate inference
- texture images
- texture classification
- feature space
- multiscale