Parameter Estimation and Segmentation of Noisy or Textured Images using the EM Algorithm and MPM Estimation.
Mary L. ComerEdward J. DelpPublished in: ICIP (2) (1994)
Keyphrases
- parameter estimation
- textured images
- em algorithm
- random fields
- unsupervised segmentation
- expectation maximization
- markov random field
- maximum likelihood
- markov fields
- maximum likelihood estimation
- texture segmentation
- mixture model
- generative model
- image segmentation
- likelihood function
- color images
- gaussian mixture model
- natural images
- maximum a posteriori
- approximate inference
- texture features
- maximum likelihood estimates
- multiresolution
- unsupervised image segmentation
- least squares
- probabilistic model
- texture information
- bayesian model selection
- energy function
- texture images
- probability density function
- segmentation algorithm
- graph cuts
- image features
- multiscale