Directional Total Generalized Variation Regularization for Impulse Noise Removal

Rasmus Dalgas Kongskov; Yiqiu Dong

doi:10.1007/978-3-319-58771-4_18

Directional Total Generalized Variation Regularization for Impulse Noise Removal

Rasmus Dalgas Kongskov, Yiqiu Dong

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Abstract

A recently suggested regularization method, which combines directional information with total generalized variation (TGV), has been shown to be successful for restoring Gaussian noise corrupted images. We extend the use of this regularizer to impulse noise removal and demonstrate that using this regularizer for directional images is highly advantageous. In order to estimate directions in impulse noise corrupted images, which is much more challenging compared to Gaussian noise corrupted images, we introduce a new Fourier transform-based method. Numerical experiments show that this method is more robust with respect to noise and also more efficient than other direction estimation methods.

Original language	English
Title of host publication	Scale Space and Variational Methods in Computer Vision
Publisher	Springer
Publication date	2017
Pages	221-231
DOIs	https://doi.org/10.1007/978-3-319-58771-4_18
Publication status	Published - 2017
Event	6^th International Conference on Scale Space and Variational Methods in Computer Vision - Hotel Koldingfjord, Kolding, Denmark Duration: 4 Jun 2017 → 8 Jun 2017 Conference number: 6

Conference

Conference	6^th International Conference on Scale Space and Variational Methods in Computer Vision
Number	6
Location	Hotel Koldingfjord
Country/Territory	Denmark
City	Kolding
Period	04/06/2017 → 08/06/2017

Series	Lecture Notes in Computer Science
Volume	10302
ISSN	0302-9743

Keywords

Directional total generalized variation
Impulse noise
Variational methods
Regularization
Image restoration

Access to Document

10.1007/978-3-319-58771-4_18

OpenUrl availability

Full text

Cite this

@inproceedings{769ac6a64f014cfe8ba2344b56c38b60,

title = "Directional Total Generalized Variation Regularization for Impulse Noise Removal",

abstract = "A recently suggested regularization method, which combines directional information with total generalized variation (TGV), has been shown to be successful for restoring Gaussian noise corrupted images. We extend the use of this regularizer to impulse noise removal and demonstrate that using this regularizer for directional images is highly advantageous. In order to estimate directions in impulse noise corrupted images, which is much more challenging compared to Gaussian noise corrupted images, we introduce a new Fourier transform-based method. Numerical experiments show that this method is more robust with respect to noise and also more efficient than other direction estimation methods.",

keywords = "Directional total generalized variation, Impulse noise, Variational methods, Regularization, Image restoration",

author = "Kongskov, \{Rasmus Dalgas\} and Yiqiu Dong",

year = "2017",

doi = "10.1007/978-3-319-58771-4\_18",

language = "English",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "221--231",

booktitle = "Scale Space and Variational Methods in Computer Vision",

note = "6<sup>th</sup> International Conference on Scale Space and Variational Methods in Computer Vision, SSVM 2017 ; Conference date: 04-06-2017 Through 08-06-2017",

}

Kongskov, RD & Dong, Y 2017, Directional Total Generalized Variation Regularization for Impulse Noise Removal. in Scale Space and Variational Methods in Computer Vision. Springer, Lecture Notes in Computer Science, vol. 10302, pp. 221-231, 6^th International Conference on Scale Space and Variational Methods in Computer Vision, Kolding, Denmark, 04/06/2017. https://doi.org/10.1007/978-3-319-58771-4_18

Directional Total Generalized Variation Regularization for Impulse Noise Removal. / Kongskov, Rasmus Dalgas; Dong, Yiqiu.
Scale Space and Variational Methods in Computer Vision. Springer, 2017. p. 221-231 (Lecture Notes in Computer Science, Vol. 10302).

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - Directional Total Generalized Variation Regularization for Impulse Noise Removal

AU - Kongskov, Rasmus Dalgas

AU - Dong, Yiqiu

N1 - Conference code: 6

PY - 2017

Y1 - 2017

N2 - A recently suggested regularization method, which combines directional information with total generalized variation (TGV), has been shown to be successful for restoring Gaussian noise corrupted images. We extend the use of this regularizer to impulse noise removal and demonstrate that using this regularizer for directional images is highly advantageous. In order to estimate directions in impulse noise corrupted images, which is much more challenging compared to Gaussian noise corrupted images, we introduce a new Fourier transform-based method. Numerical experiments show that this method is more robust with respect to noise and also more efficient than other direction estimation methods.

AB - A recently suggested regularization method, which combines directional information with total generalized variation (TGV), has been shown to be successful for restoring Gaussian noise corrupted images. We extend the use of this regularizer to impulse noise removal and demonstrate that using this regularizer for directional images is highly advantageous. In order to estimate directions in impulse noise corrupted images, which is much more challenging compared to Gaussian noise corrupted images, we introduce a new Fourier transform-based method. Numerical experiments show that this method is more robust with respect to noise and also more efficient than other direction estimation methods.

KW - Directional total generalized variation

KW - Impulse noise

KW - Variational methods

KW - Regularization

KW - Image restoration

U2 - 10.1007/978-3-319-58771-4_18

DO - 10.1007/978-3-319-58771-4_18

M3 - Article in proceedings

T3 - Lecture Notes in Computer Science

SP - 221

EP - 231

BT - Scale Space and Variational Methods in Computer Vision

PB - Springer

T2 - 6<sup>th</sup> International Conference on Scale Space and Variational Methods in Computer Vision

Y2 - 4 June 2017 through 8 June 2017

ER -

Directional Total Generalized Variation Regularization for Impulse Noise Removal

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this