Variational Bayesian inference image restoration using a product of total variation-like image priors

Giannis Chantas; Nikolaos Galatsanos; Rafael Molina; Aggelos K Katsaggelos

doi:10.1109/CIP.2010.5604259

Variational Bayesian inference image restoration using a product of total variation-like image priors

Giannis Chantas^*, Nikolaos Galatsanos, Rafael Molina, Aggelos K Katsaggelos

^*Corresponding author for this work

Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

1 Scopus citations

Abstract

In this paper a new image prior is introduced and used in image restoration. This prior is based on products of spatially weighted Total Variations (TV). These spatial weights provide this prior with the flexibilit to better capture local image features than previous TV based priors. Bayesian inference is used for image restoration with this prior via the variational approximation. The proposed algorithm is fully automatic in the sense that all necessary parameters are estimated from the data. Numerical experiments are shown which demonstrate that image restoration based on this prior compares favorably with previous state-of-the-art restoration algorithms.

Original language	English (US)
Title of host publication	2010 2nd International Workshop on Cognitive Information Processing, CIP2010
Pages	227-231
Number of pages	5
DOIs	https://doi.org/10.1109/CIP.2010.5604259
State	Published - Nov 22 2010
Event	2010 2nd International Workshop on Cognitive Information Processing, CIP2010 - Elba Island, Italy Duration: Jun 14 2010 → Jun 16 2010

Other

Other	2010 2nd International Workshop on Cognitive Information Processing, CIP2010
Country/Territory	Italy
City	Elba Island
Period	6/14/10 → 6/16/10

ASJC Scopus subject areas

Artificial Intelligence
Information Systems

Access to Document

10.1109/CIP.2010.5604259

Cite this

@inproceedings{47a164b8bd72486b867c236527818f3e,

title = "Variational Bayesian inference image restoration using a product of total variation-like image priors",

abstract = "In this paper a new image prior is introduced and used in image restoration. This prior is based on products of spatially weighted Total Variations (TV). These spatial weights provide this prior with the flexibilit to better capture local image features than previous TV based priors. Bayesian inference is used for image restoration with this prior via the variational approximation. The proposed algorithm is fully automatic in the sense that all necessary parameters are estimated from the data. Numerical experiments are shown which demonstrate that image restoration based on this prior compares favorably with previous state-of-the-art restoration algorithms.",

author = "Giannis Chantas and Nikolaos Galatsanos and Rafael Molina and Katsaggelos, {Aggelos K}",

year = "2010",

month = nov,

day = "22",

doi = "10.1109/CIP.2010.5604259",

language = "English (US)",

isbn = "9781424464593",

pages = "227--231",

booktitle = "2010 2nd International Workshop on Cognitive Information Processing, CIP2010",

note = "2010 2nd International Workshop on Cognitive Information Processing, CIP2010 ; Conference date: 14-06-2010 Through 16-06-2010",

}

Chantas, G, Galatsanos, N, Molina, R & Katsaggelos, AK 2010, Variational Bayesian inference image restoration using a product of total variation-like image priors. in 2010 2nd International Workshop on Cognitive Information Processing, CIP2010., 5604259, pp. 227-231, 2010 2nd International Workshop on Cognitive Information Processing, CIP2010, Elba Island, Italy, 6/14/10. https://doi.org/10.1109/CIP.2010.5604259

TY - GEN

T1 - Variational Bayesian inference image restoration using a product of total variation-like image priors

AU - Chantas, Giannis

AU - Galatsanos, Nikolaos

AU - Molina, Rafael

AU - Katsaggelos, Aggelos K

PY - 2010/11/22

Y1 - 2010/11/22

N2 - In this paper a new image prior is introduced and used in image restoration. This prior is based on products of spatially weighted Total Variations (TV). These spatial weights provide this prior with the flexibilit to better capture local image features than previous TV based priors. Bayesian inference is used for image restoration with this prior via the variational approximation. The proposed algorithm is fully automatic in the sense that all necessary parameters are estimated from the data. Numerical experiments are shown which demonstrate that image restoration based on this prior compares favorably with previous state-of-the-art restoration algorithms.

AB - In this paper a new image prior is introduced and used in image restoration. This prior is based on products of spatially weighted Total Variations (TV). These spatial weights provide this prior with the flexibilit to better capture local image features than previous TV based priors. Bayesian inference is used for image restoration with this prior via the variational approximation. The proposed algorithm is fully automatic in the sense that all necessary parameters are estimated from the data. Numerical experiments are shown which demonstrate that image restoration based on this prior compares favorably with previous state-of-the-art restoration algorithms.

UR - http://www.scopus.com/inward/record.url?scp=78349252114&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78349252114&partnerID=8YFLogxK

U2 - 10.1109/CIP.2010.5604259

DO - 10.1109/CIP.2010.5604259

M3 - Conference contribution

AN - SCOPUS:78349252114

SN - 9781424464593

SP - 227

EP - 231

BT - 2010 2nd International Workshop on Cognitive Information Processing, CIP2010

T2 - 2010 2nd International Workshop on Cognitive Information Processing, CIP2010

Y2 - 14 June 2010 through 16 June 2010

ER -

Variational Bayesian inference image restoration using a product of total variation-like image priors

Abstract

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this