A variational approach for restoring images corrupted by noisy blur kernels and additive noise.
Michael K. NgWei WangXile ZhaoPublished in: Numer. Linear Algebra Appl. (2017)
Keyphrases
- additive noise
- image data
- noisy images
- blind deconvolution
- spatial domain
- camera shake
- pixel wise
- image features
- multiple images
- ground truth
- three dimensional
- blur kernel
- edge detection
- test images
- object recognition
- speech signal
- map estimation
- point spread function
- speech recognition
- maximum likelihood
- motion blur
- multiresolution
- pattern recognition
- multiscale
- computer vision