Potts model parameter estimation in Bayesian segmentation of piecewise constant images.
Roxana-Gabriela RosuJean-François GiovannelliAudrey GiremusCornelia Paula VacarPublished in: ICASSP (2015)
Keyphrases
- parameter estimation
- piecewise constant
- markov fields
- maximum likelihood
- random fields
- markov random field
- segmentation algorithm
- segmentation method
- level set
- posterior distribution
- least squares
- mrf model
- image analysis
- model selection
- input image
- textured images
- statistical models
- edge detection
- level set method
- three dimensional
- em algorithm
- expectation maximization
- image classification
- image segmentation
- pixel level
- image processing
- higher order
- bayesian estimation
- natural images
- maximum entropy
- graphical models
- texture segmentation
- multiscale
- bayesian networks
- segmentation result