Middle-frequency based refinement for image super-resolution

Jae Hee Jun, Ji Hoon Choi, Jong-Ok Kim

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

This letter proposes a novel post-processing method for self-similarity based super-resolution (SR). Existing back-projection (BP) methods enhance SR images by refining the reconstructed coarse highfrequency (HF) information. However, it causes artifacts due to interpolation and excessively smoothes small HF signals, particularly in texture regions. Motivated by these observations, we propose a novel postprocessing method referred to as middle-frequency (MF) based refinement. The proposed method refines the reconstructed HF information in the MF domain rather than in the spatial domain, as in BP. In addition, it does not require an internal interpolation process, so it is free from the side-effects of interpolation. Experimental results show that the proposed algorithm provides superior performance in terms of both the quantity of reproduced HF information and the visual quality.

Original languageEnglish
Pages (from-to)300-304
Number of pages5
JournalIEICE Transactions on Information and Systems
VolumeE99D
Issue number1
DOIs
Publication statusPublished - 2016 Jan 1

Fingerprint

Interpolation
Image resolution
Refining
Textures
Processing

Keywords

  • Back-projection
  • Middle-frequency
  • Post processing
  • Superresolution

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Software
  • Artificial Intelligence
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition

Cite this

Middle-frequency based refinement for image super-resolution. / Jun, Jae Hee; Choi, Ji Hoon; Kim, Jong-Ok.

In: IEICE Transactions on Information and Systems, Vol. E99D, No. 1, 01.01.2016, p. 300-304.

Research output: Contribution to journalArticle

@article{a034f5abf72649099eeee86cdd53908b,
title = "Middle-frequency based refinement for image super-resolution",
abstract = "This letter proposes a novel post-processing method for self-similarity based super-resolution (SR). Existing back-projection (BP) methods enhance SR images by refining the reconstructed coarse highfrequency (HF) information. However, it causes artifacts due to interpolation and excessively smoothes small HF signals, particularly in texture regions. Motivated by these observations, we propose a novel postprocessing method referred to as middle-frequency (MF) based refinement. The proposed method refines the reconstructed HF information in the MF domain rather than in the spatial domain, as in BP. In addition, it does not require an internal interpolation process, so it is free from the side-effects of interpolation. Experimental results show that the proposed algorithm provides superior performance in terms of both the quantity of reproduced HF information and the visual quality.",
keywords = "Back-projection, Middle-frequency, Post processing, Superresolution",
author = "Jun, {Jae Hee} and Choi, {Ji Hoon} and Jong-Ok Kim",
year = "2016",
month = "1",
day = "1",
doi = "10.1587/transinf.2015EDL8180",
language = "English",
volume = "E99D",
pages = "300--304",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "1",

}

TY - JOUR

T1 - Middle-frequency based refinement for image super-resolution

AU - Jun, Jae Hee

AU - Choi, Ji Hoon

AU - Kim, Jong-Ok

PY - 2016/1/1

Y1 - 2016/1/1

N2 - This letter proposes a novel post-processing method for self-similarity based super-resolution (SR). Existing back-projection (BP) methods enhance SR images by refining the reconstructed coarse highfrequency (HF) information. However, it causes artifacts due to interpolation and excessively smoothes small HF signals, particularly in texture regions. Motivated by these observations, we propose a novel postprocessing method referred to as middle-frequency (MF) based refinement. The proposed method refines the reconstructed HF information in the MF domain rather than in the spatial domain, as in BP. In addition, it does not require an internal interpolation process, so it is free from the side-effects of interpolation. Experimental results show that the proposed algorithm provides superior performance in terms of both the quantity of reproduced HF information and the visual quality.

AB - This letter proposes a novel post-processing method for self-similarity based super-resolution (SR). Existing back-projection (BP) methods enhance SR images by refining the reconstructed coarse highfrequency (HF) information. However, it causes artifacts due to interpolation and excessively smoothes small HF signals, particularly in texture regions. Motivated by these observations, we propose a novel postprocessing method referred to as middle-frequency (MF) based refinement. The proposed method refines the reconstructed HF information in the MF domain rather than in the spatial domain, as in BP. In addition, it does not require an internal interpolation process, so it is free from the side-effects of interpolation. Experimental results show that the proposed algorithm provides superior performance in terms of both the quantity of reproduced HF information and the visual quality.

KW - Back-projection

KW - Middle-frequency

KW - Post processing

KW - Superresolution

UR - http://www.scopus.com/inward/record.url?scp=84953326281&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84953326281&partnerID=8YFLogxK

U2 - 10.1587/transinf.2015EDL8180

DO - 10.1587/transinf.2015EDL8180

M3 - Article

AN - SCOPUS:84953326281

VL - E99D

SP - 300

EP - 304

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 1

ER -