A comparative study on the K-views classifier and Markov random fields for image texture classification.
Minh PhamMei XiangChih-Cheng HungBor-Chen KuoPublished in: ACM Southeast Regional Conference (1) (2005)
Keyphrases
- texture classification
- markov random field
- textured images
- random fields
- image segmentation
- energy function
- image analysis
- low level vision
- texture images
- texture features
- mrf model
- natural textures
- image features
- texture analysis
- graph cuts
- highly discriminative
- texture segmentation
- input image
- higher order
- multiscale
- maximum a posteriori
- unsupervised segmentation
- parameter estimation
- image restoration
- piecewise constant functions
- training data
- conditional random fields
- image classification
- local binary pattern
- texture model
- feature selection
- texture information
- color images
- pairwise
- similarity measure
- image matching
- texture descriptors
- spatial information
- post processing
- training set
- object recognition
- computer vision