Dynamic programming approach for context classification using the Markov random field.
Robert M. HaralickMing Chua ZhangRoger W. EhrichPublished in: ICPR (1988)
Keyphrases
- markov random field
- dynamic programming
- mrf model
- image restoration
- higher order
- energy function
- discriminative random fields
- graph cuts
- random fields
- image segmentation
- maximum a posteriori
- parameter estimation
- belief propagation
- energy minimization
- potential functions
- map estimation
- machine learning
- contextual constraints
- loopy belief propagation
- prior model
- feature vectors
- contextual information
- support vector
- feature selection
- image classification
- training data
- mrf models
- markov random field model
- state space
- denoising