Blind deconvolution for single noisy and blurry image using alternating maximum a posteriori estimation with low rank prior.
Shijie SunHuaici ZhaoJinfeng LvMingguo HaoBo LiPublished in: ISCIT (2016)
Keyphrases
- blind deconvolution
- low rank
- maximum a posteriori estimation
- total variation
- convex optimization
- blur kernel
- matrix factorization
- missing data
- image deblurring
- multi frame
- blurred images
- point spread function
- map estimation
- bayesian framework
- singular value decomposition
- single image
- noise reduction
- linear combination
- input image
- semi supervised
- multiscale
- image processing
- minimization problems
- image restoration
- dimensionality reduction
- image data
- prior knowledge
- bayesian networks
- similarity measure