Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Automated parameter selection in the L1-L2-TV model for removing Gaussian plus impulse noise

Langer, Andreas LU (2017) In Inverse Problems 33(7).
Abstract

The minimization of a functional consisting of a combined L 1/L 2-data-fidelity term and a total variation term, named L 1-L 2-TV model, is considered to remove a mixture of Gaussian and impulse noise in images, which are possibly additionally deformed by some convolution operator. We investigate analytically the stability of this model with respect to its parameters and link it to a constrained minimization problem. Based on these investigations and a statistical characterization of the mixed Gaussian-impulse noise a fully automated parameter selection algorithm for the L 1-L 2-TV model is presented. It is shown by numerical experiments that the proposed method finds... (More)

The minimization of a functional consisting of a combined L 1/L 2-data-fidelity term and a total variation term, named L 1-L 2-TV model, is considered to remove a mixture of Gaussian and impulse noise in images, which are possibly additionally deformed by some convolution operator. We investigate analytically the stability of this model with respect to its parameters and link it to a constrained minimization problem. Based on these investigations and a statistical characterization of the mixed Gaussian-impulse noise a fully automated parameter selection algorithm for the L 1-L 2-TV model is presented. It is shown by numerical experiments that the proposed method finds parameters with which noise is removed considerably while features are preserved in images.

(Less)
Please use this url to cite or link to this publication:
author
publishing date
type
Contribution to journal
publication status
published
subject
keywords
constrained/unconstrained problem, image reconstruction, mixed noise, parameter selection, total variation minimization
in
Inverse Problems
volume
33
issue
7
article number
074002
publisher
IOP Publishing
external identifiers
  • scopus:85021757409
ISSN
0266-5611
DOI
10.1088/1361-6420/33/7/074002
language
English
LU publication?
no
additional info
Publisher Copyright: © 2017 IOP Publishing Ltd. Copyright: Copyright 2017 Elsevier B.V., All rights reserved.
id
81c2c16c-77a0-409c-bb9b-4408518c5926
date added to LUP
2021-03-15 22:30:31
date last changed
2022-03-26 18:40:18
@article{81c2c16c-77a0-409c-bb9b-4408518c5926,
  abstract     = {{<p>The minimization of a functional consisting of a combined L <sup>1</sup>/L <sup>2</sup>-data-fidelity term and a total variation term, named L <sup>1</sup>-L <sup>2</sup>-TV model, is considered to remove a mixture of Gaussian and impulse noise in images, which are possibly additionally deformed by some convolution operator. We investigate analytically the stability of this model with respect to its parameters and link it to a constrained minimization problem. Based on these investigations and a statistical characterization of the mixed Gaussian-impulse noise a fully automated parameter selection algorithm for the L <sup>1</sup>-L <sup>2</sup>-TV model is presented. It is shown by numerical experiments that the proposed method finds parameters with which noise is removed considerably while features are preserved in images.</p>}},
  author       = {{Langer, Andreas}},
  issn         = {{0266-5611}},
  keywords     = {{constrained/unconstrained problem; image reconstruction; mixed noise; parameter selection; total variation minimization}},
  language     = {{eng}},
  month        = {{06}},
  number       = {{7}},
  publisher    = {{IOP Publishing}},
  series       = {{Inverse Problems}},
  title        = {{Automated parameter selection in the L<sup>1</sup>-L<sup>2</sup>-TV model for removing Gaussian plus impulse noise}},
  url          = {{http://dx.doi.org/10.1088/1361-6420/33/7/074002}},
  doi          = {{10.1088/1361-6420/33/7/074002}},
  volume       = {{33}},
  year         = {{2017}},
}