Contextual remote-sensing image classification through support vector machines, Markov random fields and graph cuts.
Andrea De GiorgiGabriele MoserSebastiano Bruno SerpicoPublished in: IGARSS (2014)
Keyphrases
- remote sensing
- graph cuts
- markov random field
- image classification
- support vector
- remotely sensed data
- belief propagation
- multi label
- energy minimization
- multispectral
- energy function
- image analysis
- image segmentation
- higher order
- image processing
- parameter estimation
- mrf model
- maximum a posteriori
- object segmentation
- remote sensing data
- random fields
- image restoration
- message passing
- high resolution
- conditional random fields
- image features
- potential functions
- map inference
- feature extraction
- pairwise
- loopy belief propagation
- mrf optimization
- interactive segmentation
- low level vision
- visual words
- segmentation method
- maximum likelihood
- high quality
- min cut
- iterated conditional modes