Single image super resolution by ℓ2 approximation with random sampled dictionary

Takanori Fujisawa, Taichi Yoshida, Kazu Mishiba, Masaaki Ikehara

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

In this paper, we propose an example-based single image super resolution (SR) method by ℓ2 approximation with self-sampled image patches. Example-based super resolution methods can reconstruct high resolution image patches by a linear combination of atoms in an overcomplete dictionary. This reconstruction requires a pair of two dictionaries created by tremendous low and high resolution image pairs from the prepared image databases. In our method, we introduce the dictionary by random sampling patches from just an input image and eliminate its training process. This dictionary exploits the self-similarity of images and it will no more depend on external image sets, which consern the storage space or the accuracy of referred image sets. In addition, we modified the approximation of input image to an ℓ2-norm minimization problem, instead of commonly used sparse approximation such as.1-norm regularization. The ℓ2 approximation has an advantage of computational cost by only solving an inverse problem. Through some experiments, the proposed method drastically reduces the computational time for the SR, and it provides a comparable performance to the conventional example-based SR methods with an.1 approximation and dictionary training.

Original languageEnglish
Pages (from-to)612-620
Number of pages9
JournalIEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
VolumeE99A
Issue number2
DOIs
Publication statusPublished - 2016 Feb 1

Fingerprint

Super-resolution
Glossaries
Approximation
Image resolution
Patch
Optical resolving power
Inverse problems
High Resolution
Sparse Approximation
Norm
Dictionary
Sampling
Random Sampling
Atoms
Image Database
Self-similarity
Minimization Problem
Linear Combination
Computational Cost
Regularization

Keywords

  • Examplebased
  • Single image super resolution
  • ℓ-norm minimization

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Graphics and Computer-Aided Design
  • Applied Mathematics
  • Signal Processing

Cite this

Single image super resolution by ℓ2 approximation with random sampled dictionary. / Fujisawa, Takanori; Yoshida, Taichi; Mishiba, Kazu; Ikehara, Masaaki.

In: IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, Vol. E99A, No. 2, 01.02.2016, p. 612-620.

Research output: Contribution to journalArticle

@article{f568d96bae8d463abcecbdea4192cffa,
title = "Single image super resolution by ℓ2 approximation with random sampled dictionary",
abstract = "In this paper, we propose an example-based single image super resolution (SR) method by ℓ2 approximation with self-sampled image patches. Example-based super resolution methods can reconstruct high resolution image patches by a linear combination of atoms in an overcomplete dictionary. This reconstruction requires a pair of two dictionaries created by tremendous low and high resolution image pairs from the prepared image databases. In our method, we introduce the dictionary by random sampling patches from just an input image and eliminate its training process. This dictionary exploits the self-similarity of images and it will no more depend on external image sets, which consern the storage space or the accuracy of referred image sets. In addition, we modified the approximation of input image to an ℓ2-norm minimization problem, instead of commonly used sparse approximation such as.1-norm regularization. The ℓ2 approximation has an advantage of computational cost by only solving an inverse problem. Through some experiments, the proposed method drastically reduces the computational time for the SR, and it provides a comparable performance to the conventional example-based SR methods with an.1 approximation and dictionary training.",
keywords = "Examplebased, Single image super resolution, ℓ-norm minimization",
author = "Takanori Fujisawa and Taichi Yoshida and Kazu Mishiba and Masaaki Ikehara",
year = "2016",
month = "2",
day = "1",
doi = "10.1587/transfun.E99.A.612",
language = "English",
volume = "E99A",
pages = "612--620",
journal = "IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences",
issn = "0916-8508",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "2",

}

TY - JOUR

T1 - Single image super resolution by ℓ2 approximation with random sampled dictionary

AU - Fujisawa, Takanori

AU - Yoshida, Taichi

AU - Mishiba, Kazu

AU - Ikehara, Masaaki

PY - 2016/2/1

Y1 - 2016/2/1

N2 - In this paper, we propose an example-based single image super resolution (SR) method by ℓ2 approximation with self-sampled image patches. Example-based super resolution methods can reconstruct high resolution image patches by a linear combination of atoms in an overcomplete dictionary. This reconstruction requires a pair of two dictionaries created by tremendous low and high resolution image pairs from the prepared image databases. In our method, we introduce the dictionary by random sampling patches from just an input image and eliminate its training process. This dictionary exploits the self-similarity of images and it will no more depend on external image sets, which consern the storage space or the accuracy of referred image sets. In addition, we modified the approximation of input image to an ℓ2-norm minimization problem, instead of commonly used sparse approximation such as.1-norm regularization. The ℓ2 approximation has an advantage of computational cost by only solving an inverse problem. Through some experiments, the proposed method drastically reduces the computational time for the SR, and it provides a comparable performance to the conventional example-based SR methods with an.1 approximation and dictionary training.

AB - In this paper, we propose an example-based single image super resolution (SR) method by ℓ2 approximation with self-sampled image patches. Example-based super resolution methods can reconstruct high resolution image patches by a linear combination of atoms in an overcomplete dictionary. This reconstruction requires a pair of two dictionaries created by tremendous low and high resolution image pairs from the prepared image databases. In our method, we introduce the dictionary by random sampling patches from just an input image and eliminate its training process. This dictionary exploits the self-similarity of images and it will no more depend on external image sets, which consern the storage space or the accuracy of referred image sets. In addition, we modified the approximation of input image to an ℓ2-norm minimization problem, instead of commonly used sparse approximation such as.1-norm regularization. The ℓ2 approximation has an advantage of computational cost by only solving an inverse problem. Through some experiments, the proposed method drastically reduces the computational time for the SR, and it provides a comparable performance to the conventional example-based SR methods with an.1 approximation and dictionary training.

KW - Examplebased

KW - Single image super resolution

KW - ℓ-norm minimization

UR - http://www.scopus.com/inward/record.url?scp=84957657916&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84957657916&partnerID=8YFLogxK

U2 - 10.1587/transfun.E99.A.612

DO - 10.1587/transfun.E99.A.612

M3 - Article

VL - E99A

SP - 612

EP - 620

JO - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

JF - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

SN - 0916-8508

IS - 2

ER -