Spectral-spatial classification based on subspace support vector machine and Markov random field.
Haoyang YuLianru GaoJun LiBing ZhangPublished in: IGARSS (2016)
Keyphrases
- markov random field
- support vector machine
- feature space
- energy function
- svm classifier
- graph cuts
- image segmentation
- pairwise
- energy minimization
- support vector
- discriminative random fields
- belief propagation
- random fields
- higher order
- feature vectors
- parameter estimation
- mrf model
- maximum a posteriori
- feature extraction
- image restoration
- feature selection
- machine learning
- training set
- image classification
- potential functions
- spectral signatures
- prior model
- map estimation
- multi class
- message passing
- neighboring pixels
- active contours
- dimensionality reduction
- simulated annealing
- conditional random fields
- markov random field model
- higher order cliques
- spatial continuity
- smoothness prior
- maximum a posteriori probability
- mrf models
- model selection