Expectation-Maximization with Image-Weighted Markov Random Fields to Handle Severe Pathology.
Alex M. PagnozziNicholas D. H. DowsonAndrew P. BradleyRoslyn N. BoydPierrick BourgeatStephen E. RosePublished in: DICTA (2015)
Keyphrases
- markov random field
- random fields
- image segmentation
- low level vision
- mrf model
- energy function
- maximum a posteriori
- expectation maximization
- map estimation
- neighboring pixels
- prior model
- parameter estimation
- textured images
- markov mesh random field
- prior models
- low level vision tasks
- image analysis
- graph cuts
- higher order
- input image
- mrf models
- image features
- image restoration
- em algorithm
- piecewise constant functions
- conditional random fields
- belief propagation
- multiscale
- energy minimization
- potential functions
- bayesian framework
- maximum a posteriori probability
- label field
- pairwise
- segmentation method
- probability density function
- image understanding
- maximum likelihood
- non stationary
- natural images
- texture model
- similarity measure
- smoothness prior
- mixture model
- image processing
- active contours
- loopy belief propagation
- segmentation algorithm
- high resolution
- edge detection