Learning from Ideal Edge for Image Restoration

Jin-Ping HE; Kun GAO; Guo-Qiang NI; Guang-Da SU; Jian-Sheng CHEN

doi:10.1587/transinf.E96.D.2487

Learning from Ideal Edge for Image Restoration

Jin-Ping HE, Kun GAO, Guo-Qiang NI, Guang-Da SU, Jian-Sheng CHEN

Full Text Views

0

Share
Cite this

Summary :

Considering the real existent fact of the ideal edge and the learning style of image analogy without reference parameters, a blind image recovery algorithm using a self-adaptive learning method is proposed in this paper. We show that a specific local image patch with degradation characteristic can be utilized for restoring the whole image. In the training process, a clear counterpart of the local image patch is constructed based on the ideal edge assumption so that identification of the Point Spread Function is no longer needed. Experiments demonstrate the effectiveness of the proposed method on remote sensing images.

Publication: IEICE TRANSACTIONS on Information Vol.E96-D No.11 pp.2487-2491

Publication Date: 2013/11/01

Publicized

Online ISSN: 1745-1361

DOI: 10.1587/transinf.E96.D.2487

Type of Manuscript: LETTER

Category: Image Processing and Video Processing

Authors

Jin-Ping HE
  Beijing Institute of Space Mechanics & Electricity
Kun GAO
  School of Optoelectronics, Beijing Institute of Technology
Guo-Qiang NI
  School of Optoelectronics, Beijing Institute of Technology
Guang-Da SU
  Tsinghua University
Jian-Sheng CHEN
  Tsinghua University

Keyword

learning-based, image restoration, ideal edge, image analogy

Cite this

Copy

Jin-Ping HE, Kun GAO, Guo-Qiang NI, Guang-Da SU, Jian-Sheng CHEN, "Learning from Ideal Edge for Image Restoration" in IEICE TRANSACTIONS on Information, vol. E96-D, no. 11, pp. 2487-2491, November 2013, doi: 10.1587/transinf.E96.D.2487.
Abstract: Considering the real existent fact of the ideal edge and the learning style of image analogy without reference parameters, a blind image recovery algorithm using a self-adaptive learning method is proposed in this paper. We show that a specific local image patch with degradation characteristic can be utilized for restoring the whole image. In the training process, a clear counterpart of the local image patch is constructed based on the ideal edge assumption so that identification of the Point Spread Function is no longer needed. Experiments demonstrate the effectiveness of the proposed method on remote sensing images.
URL: https://globals.ieice.org/en_transactions/information/10.1587/transinf.E96.D.2487/_p

Copy

@ARTICLE{e96-d_11_2487,
author={Jin-Ping HE, Kun GAO, Guo-Qiang NI, Guang-Da SU, Jian-Sheng CHEN, },
journal={IEICE TRANSACTIONS on Information},
title={Learning from Ideal Edge for Image Restoration},
year={2013},
volume={E96-D},
number={11},
pages={2487-2491},
abstract={Considering the real existent fact of the ideal edge and the learning style of image analogy without reference parameters, a blind image recovery algorithm using a self-adaptive learning method is proposed in this paper. We show that a specific local image patch with degradation characteristic can be utilized for restoring the whole image. In the training process, a clear counterpart of the local image patch is constructed based on the ideal edge assumption so that identification of the Point Spread Function is no longer needed. Experiments demonstrate the effectiveness of the proposed method on remote sensing images.},
keywords={},
doi={10.1587/transinf.E96.D.2487},
ISSN={1745-1361},
month={November},}

Copy

TY - JOUR
TI - Learning from Ideal Edge for Image Restoration
T2 - IEICE TRANSACTIONS on Information
SP - 2487
EP - 2491
AU - Jin-Ping HE
AU - Kun GAO
AU - Guo-Qiang NI
AU - Guang-Da SU
AU - Jian-Sheng CHEN
PY - 2013
DO - 10.1587/transinf.E96.D.2487
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E96-D
IS - 11
JA - IEICE TRANSACTIONS on Information
Y1 - November 2013
AB - Considering the real existent fact of the ideal edge and the learning style of image analogy without reference parameters, a blind image recovery algorithm using a self-adaptive learning method is proposed in this paper. We show that a specific local image patch with degradation characteristic can be utilized for restoring the whole image. In the training process, a clear counterpart of the local image patch is constructed based on the ideal edge assumption so that identification of the Point Spread Function is no longer needed. Experiments demonstrate the effectiveness of the proposed method on remote sensing images.
ER -