A discrete-time recurrent neural network for image restoration from its Fourier spectrum magnitudes.
Adrian BurianJukka SaarinenPauli KuosmanenPublished in: Int. J. Comput. Syst. Signals (2002)
Keyphrases
- recurrent neural networks
- image restoration
- fourier spectrum
- edge preserving
- feed forward
- total variation
- neural network
- markov random field
- artificial neural networks
- iterative image restoration
- regularization parameter
- recurrent networks
- super resolution
- blurred images
- image processing
- neural model
- modified hopfield neural network
- wiener filter
- motion blur
- regularization method
- low level vision
- image deconvolution
- image deblurring
- degraded images
- restored image
- point spread function
- gaussian noise
- restoration algorithm
- blur identification
- image restoration problems
- pairwise