A segmentation method for textured images based on the maximum posterior mode criterion.
Frédéric LehmannPublished in: ICASSP (2011)
Keyphrases
- segmentation method
- textured images
- texture segmentation
- unsupervised segmentation
- image segmentation
- markov random field
- random fields
- active contour model
- color images
- natural images
- energy function
- texture features
- texture images
- segmentation algorithm
- region growing
- active contours
- object segmentation
- texture information
- conditional random fields
- superpixels
- region merging
- probabilistic model
- segmented images
- segmentation result
- shape prior
- image processing
- initial segmentation
- machine learning
- maximum entropy
- bayesian framework
- level set
- higher order