A new stochastic image model based on Markov random fields and its application to texture modeling.
Siamak YousefiNasser KehtarnavazPublished in: ICASSP (2011)
Keyphrases
- markov random field
- image segmentation
- random fields
- energy function
- mrf model
- low level vision
- neighboring pixels
- map estimation
- prior model
- prior models
- maximum a posteriori probability
- belief propagation
- textured images
- graph cuts
- parameter estimation
- energy minimization
- mrf models
- maximum a posterior
- maximum a posteriori
- input image
- higher order
- image analysis
- piecewise constant functions
- segmentation method
- image restoration
- low level vision tasks
- markov mesh random field
- multiscale
- image features
- edge detection
- pairwise
- label field
- gauss markov random fields
- conditional random fields
- smoothness prior
- image classification
- pixel values
- discriminative random fields
- bayesian framework
- map inference
- gray level
- non stationary
- texture model
- image prior
- prior information
- image understanding
- probability density function
- potential functions