A new inverse filter criterion for blind deconvolution of spiky signals using Gaussian mixtures.
Ignacio Santamaría-CaballeroCarlos J. Pantaleón-PrietoFernando Díaz-de-MaríaAntonio Artés-RodríguezPublished in: ICASSP (1996)
Keyphrases
- blind deconvolution
- gaussian mixture
- noise reduction
- minimum classification error
- total variation
- expectation maximization
- em algorithm
- closed form
- probability density function
- image deblurring
- covariance matrix
- multi frame
- gaussian mixture model
- mixture model
- bayesian framework
- point spread function
- map estimation
- density estimation
- image restoration
- filtering algorithm
- computer vision
- denoising
- blurred images
- feature extraction
- median filter
- low resolution