Maximum likelihood blur identification and image restoration using the EM algorithm.
Aggelos K. KatsaggelosKuen-Tsair LayPublished in: IEEE Trans. Signal Process. (1991)
Keyphrases
- blur identification
- em algorithm
- maximum likelihood
- image restoration
- expectation maximization
- mixture model
- parameter estimation
- point spread function
- blind image restoration
- blurred images
- maximum likelihood estimation
- total variation
- image processing
- gaussian mixture model
- bayesian estimation
- generative model
- maximum a posteriori
- hyperparameters
- super resolution
- motion blur
- markov random field
- additive noise
- incomplete data
- regularization parameter
- map estimation
- log likelihood function
- hidden variables
- gaussian distribution
- posterior distribution
- noisy images
- alternating minimization
- high resolution
- probability density function
- subband
- higher order
- bayesian networks