Illumination invariants based on Markov random fields.
Pavel VachaMichal HaindlPublished in: ICPR (2008)
Keyphrases
- markov random field
- maximum a posteriori
- graph cuts
- belief propagation
- parameter estimation
- pairwise
- mrf model
- higher order
- random fields
- conditional random fields
- image segmentation
- single image
- image restoration
- energy function
- illumination invariant
- textured images
- illumination conditions
- low level vision
- potential functions
- energy minimization
- potts model
- map inference
- markov networks
- image analysis
- prior models
- iterative conditional
- texture model
- map estimation
- message passing