Variational Dirichlet Blur Kernel Estimation

Variational Dirichlet Blur Kernel Estimation. Zhou, X., Mateos, J., Zhou, F., Molina, R., & Katsaggelos, A. K. IEEE Transactions on Image Processing, 24(12):5127–5139, dec, 2015.

Paper doi abstract bibtex

Blind image deconvolution involves two key objectives: 1) latent image and 2) blur estimation. For latent image estimation, we propose a fast deconvolution algorithm, which uses an image prior of nondimensional Gaussianity measure to enforce sparsity and an undetermined boundary condition methodology to reduce boundary artifacts. For blur estimation, a linear inverse problem with normalization and nonnegative constraints must be solved. However, the normalization constraint is ignored in many blind image deblurring methods, mainly because it makes the problem less tractable. In this paper, we show that the normalization constraint can be very naturally incorporated into the estimation process by using a Dirichlet distribution to approximate the posterior distribution of the blur. Making use of variational Dirichlet approximation, we provide a blur posterior approximation that considers the uncertainty of the estimate and removes noise in the estimated kernel. Experiments with synthetic and real data demonstrate that the proposed method is very competitive to the state-of-the-art blind image restoration methods.

@article{Xu2015,
abstract = {Blind image deconvolution involves two key objectives: 1) latent image and 2) blur estimation. For latent image estimation, we propose a fast deconvolution algorithm, which uses an image prior of nondimensional Gaussianity measure to enforce sparsity and an undetermined boundary condition methodology to reduce boundary artifacts. For blur estimation, a linear inverse problem with normalization and nonnegative constraints must be solved. However, the normalization constraint is ignored in many blind image deblurring methods, mainly because it makes the problem less tractable. In this paper, we show that the normalization constraint can be very naturally incorporated into the estimation process by using a Dirichlet distribution to approximate the posterior distribution of the blur. Making use of variational Dirichlet approximation, we provide a blur posterior approximation that considers the uncertainty of the estimate and removes noise in the estimated kernel. Experiments with synthetic and real data demonstrate that the proposed method is very competitive to the state-of-the-art blind image restoration methods.},
author = {Zhou, Xu and Mateos, Javier and Zhou, Fugen and Molina, Rafael and Katsaggelos, Aggelos K.},
doi = {10.1109/TIP.2015.2478407},
issn = {1057-7149},
journal = {IEEE Transactions on Image Processing},
keywords = {Blind deconvolution,Dirichlet distribution,constrained optimization,image deblurring,inverse problem,point spread function,variational distribution approximations},
month = {dec},
number = {12},
pages = {5127--5139},
title = {{Variational Dirichlet Blur Kernel Estimation}},
url = {http://ieeexplore.ieee.org/document/7265038/},
volume = {24},
year = {2015}
}

Downloads: 0

{"_id":"7YK5hT2GDvJQrJsMo","bibbaseid":"zhou-mateos-zhou-molina-katsaggelos-variationaldirichletblurkernelestimation-2015","author_short":["Zhou, X.","Mateos, J.","Zhou, F.","Molina, R.","Katsaggelos, A. K."],"bibdata":{"bibtype":"article","type":"article","abstract":"Blind image deconvolution involves two key objectives: 1) latent image and 2) blur estimation. For latent image estimation, we propose a fast deconvolution algorithm, which uses an image prior of nondimensional Gaussianity measure to enforce sparsity and an undetermined boundary condition methodology to reduce boundary artifacts. For blur estimation, a linear inverse problem with normalization and nonnegative constraints must be solved. However, the normalization constraint is ignored in many blind image deblurring methods, mainly because it makes the problem less tractable. In this paper, we show that the normalization constraint can be very naturally incorporated into the estimation process by using a Dirichlet distribution to approximate the posterior distribution of the blur. Making use of variational Dirichlet approximation, we provide a blur posterior approximation that considers the uncertainty of the estimate and removes noise in the estimated kernel. Experiments with synthetic and real data demonstrate that the proposed method is very competitive to the state-of-the-art blind image restoration methods.","author":[{"propositions":[],"lastnames":["Zhou"],"firstnames":["Xu"],"suffixes":[]},{"propositions":[],"lastnames":["Mateos"],"firstnames":["Javier"],"suffixes":[]},{"propositions":[],"lastnames":["Zhou"],"firstnames":["Fugen"],"suffixes":[]},{"propositions":[],"lastnames":["Molina"],"firstnames":["Rafael"],"suffixes":[]},{"propositions":[],"lastnames":["Katsaggelos"],"firstnames":["Aggelos","K."],"suffixes":[]}],"doi":"10.1109/TIP.2015.2478407","issn":"1057-7149","journal":"IEEE Transactions on Image Processing","keywords":"Blind deconvolution,Dirichlet distribution,constrained optimization,image deblurring,inverse problem,point spread function,variational distribution approximations","month":"dec","number":"12","pages":"5127–5139","title":"Variational Dirichlet Blur Kernel Estimation","url":"http://ieeexplore.ieee.org/document/7265038/","volume":"24","year":"2015","bibtex":"@article{Xu2015,\nabstract = {Blind image deconvolution involves two key objectives: 1) latent image and 2) blur estimation. For latent image estimation, we propose a fast deconvolution algorithm, which uses an image prior of nondimensional Gaussianity measure to enforce sparsity and an undetermined boundary condition methodology to reduce boundary artifacts. For blur estimation, a linear inverse problem with normalization and nonnegative constraints must be solved. However, the normalization constraint is ignored in many blind image deblurring methods, mainly because it makes the problem less tractable. In this paper, we show that the normalization constraint can be very naturally incorporated into the estimation process by using a Dirichlet distribution to approximate the posterior distribution of the blur. Making use of variational Dirichlet approximation, we provide a blur posterior approximation that considers the uncertainty of the estimate and removes noise in the estimated kernel. Experiments with synthetic and real data demonstrate that the proposed method is very competitive to the state-of-the-art blind image restoration methods.},\nauthor = {Zhou, Xu and Mateos, Javier and Zhou, Fugen and Molina, Rafael and Katsaggelos, Aggelos K.},\ndoi = {10.1109/TIP.2015.2478407},\nissn = {1057-7149},\njournal = {IEEE Transactions on Image Processing},\nkeywords = {Blind deconvolution,Dirichlet distribution,constrained optimization,image deblurring,inverse problem,point spread function,variational distribution approximations},\nmonth = {dec},\nnumber = {12},\npages = {5127--5139},\ntitle = {{Variational Dirichlet Blur Kernel Estimation}},\nurl = {http://ieeexplore.ieee.org/document/7265038/},\nvolume = {24},\nyear = {2015}\n}\n","author_short":["Zhou, X.","Mateos, J.","Zhou, F.","Molina, R.","Katsaggelos, A. K."],"key":"Xu2015","id":"Xu2015","bibbaseid":"zhou-mateos-zhou-molina-katsaggelos-variationaldirichletblurkernelestimation-2015","role":"author","urls":{"Paper":"http://ieeexplore.ieee.org/document/7265038/"},"keyword":["Blind deconvolution","Dirichlet distribution","constrained optimization","image deblurring","inverse problem","point spread function","variational distribution approximations"],"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://sites.northwestern.edu/ivpl/files/2023/06/IVPL_Updated_publications-1.bib","dataSources":["KTWAakbPXLGfYseXn","ePKPjG8C6yvpk4mEK","ya2CyA73rpZseyrZ8","D8k2SxfC5dKNRFgro","7Dwzbxq93HWrJEhT6","qhF8zxmGcJfvtdeAg","fvDEHD49E2ZRwE3fb","H7crv8NWhZup4d4by","DHqokWsryttGh7pJE","vRJd4wNg9HpoZSMHD","sYxQ6pxFgA59JRhxi","w2WahSbYrbcCKBDsC","XasdXLL99y5rygCmq","3gkSihZQRfAD2KBo3","t5XMbyZbtPBo4wBGS","bEpHM2CtrwW2qE8FP","teJzFLHexaz5AQW5z"],"keywords":["blind deconvolution","dirichlet distribution","constrained optimization","image deblurring","inverse problem","point spread function","variational distribution approximations"],"search_terms":["variational","dirichlet","blur","kernel","estimation","zhou","mateos","zhou","molina","katsaggelos"],"title":"Variational Dirichlet Blur Kernel Estimation","year":2015}