Gaussian mixture Markov random field for image denoising and reconstruction.
Ruoqiao ZhangCharles A. BoumanJean-Baptiste ThibaultKen D. SauerPublished in: GlobalSIP (2013)
Keyphrases
- markov random field
- image denoising
- gaussian mixture
- natural images
- denoising
- low level vision problems
- higher order
- noisy images
- covariance matrix
- gaussian mixture model
- belief propagation
- total variation
- expectation maximization
- image processing
- em algorithm
- parameter estimation
- probability density function
- maximum a posteriori
- energy function
- closed form
- random fields
- graph cuts
- density estimation
- image restoration
- energy minimization
- pairwise
- mixture model
- wavelet coefficients
- image segmentation
- textured images
- denoising algorithm
- image patches
- low level vision
- potential functions
- image reconstruction
- high resolution
- markov networks
- map estimation
- prior model
- maximum likelihood
- image prior
- similarity measure
- computational complexity