Parallel image classification using multiscale Markov random fields.
Zoltan KatoMarc BerthodJosiane ZerubiaPublished in: ICASSP (5) (1993)
Keyphrases
- markov random field
- image classification
- multiscale
- image segmentation
- image representation
- mrf model
- higher order
- belief propagation
- parameter estimation
- random fields
- energy minimization
- maximum a posteriori
- graph cuts
- image restoration
- pairwise
- energy function
- bag of words
- message passing
- low level vision
- natural images
- map inference
- image processing
- image features
- feature extraction
- superpixels
- potential functions
- partition function
- conditional random fields
- potts model
- computer vision
- image analysis
- textured images
- loopy belief propagation
- machine learning
- map estimation
- approximate inference
- image understanding
- efficient inference
- texture model
- multi label