Markov random fields for joint unmixing and segmentation of hyperspectral images.
Olivier EchesNicolas DobigeonJean-Yves TourneretPublished in: WHISPERS (2010)
Keyphrases
- hyperspectral images
- markov random field
- image segmentation
- energy function
- mrf model
- hyperspectral
- low level vision
- hyperspectral data
- textured images
- label field
- graph cuts
- hyperspectral imagery
- belief propagation
- maximum a posteriori
- remote sensing
- random fields
- segmentation algorithm
- scene labeling
- parameter estimation
- multispectral
- image restoration
- pixel classification
- energy minimization
- higher order
- level set
- pairwise
- potential functions
- discriminative random fields
- region growing
- conditional random fields
- segmentation method
- texture segmentation
- infrared
- computer vision
- multispectral images
- random projections
- image analysis
- multiscale
- image processing
- gibbs energy