Tampere University of Technology

TUTCRIS Research Portal

Image interpolation based on non-local geometric similarities

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Standard

Image interpolation based on non-local geometric similarities. / Zhu, Shuyuan; Zeng, Bing; Liu, Guanghui; Zeng, Liaoyuan; Fang, Lu; Gabbouj, Moncef.

2015 IEEE International Conference on Multimedia and Expo (ICME). IEEE COMPUTER SOCIETY PRESS, 2015.

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Harvard

Zhu, S, Zeng, B, Liu, G, Zeng, L, Fang, L & Gabbouj, M 2015, Image interpolation based on non-local geometric similarities. in 2015 IEEE International Conference on Multimedia and Expo (ICME). IEEE COMPUTER SOCIETY PRESS, IEEE International Conference on Multimedia and Expo, 1/01/00. https://doi.org/10.1109/ICME.2015.7177417

APA

Zhu, S., Zeng, B., Liu, G., Zeng, L., Fang, L., & Gabbouj, M. (2015). Image interpolation based on non-local geometric similarities. In 2015 IEEE International Conference on Multimedia and Expo (ICME) IEEE COMPUTER SOCIETY PRESS. https://doi.org/10.1109/ICME.2015.7177417

Vancouver

Zhu S, Zeng B, Liu G, Zeng L, Fang L, Gabbouj M. Image interpolation based on non-local geometric similarities. In 2015 IEEE International Conference on Multimedia and Expo (ICME). IEEE COMPUTER SOCIETY PRESS. 2015 https://doi.org/10.1109/ICME.2015.7177417

Author

Zhu, Shuyuan ; Zeng, Bing ; Liu, Guanghui ; Zeng, Liaoyuan ; Fang, Lu ; Gabbouj, Moncef. / Image interpolation based on non-local geometric similarities. 2015 IEEE International Conference on Multimedia and Expo (ICME). IEEE COMPUTER SOCIETY PRESS, 2015.

Bibtex - Download

@inproceedings{02ef04121cf04f268ae67d415bedaa25,
title = "Image interpolation based on non-local geometric similarities",
abstract = "Image interpolation refers to constructing a high-resolution (HR) image from a low-resolution (LR) image. Traditionally, an HR image can be produced from an observed LR image via the polynomial-based interpolation (bi-linear or bi-cubic interpolations, involving a small number of neighbors around each interpolated position). The advanced interpolation makes use of the so-called 'geometric similarity' to design a set of optimal interpolation weighting coefficients. However, better geometric similarities can perhaps be found from a non-local area within the LR source image or even from other but similar images (possibly with higher resolutions). Based on this fact, we propose in this paper a non-local geometric similarity based interpolation scheme to construct HR images. In our proposed method, optimal weighting coefficients are determined by solving a regularized least squares problem which is built upon a number of dual reference patches drawn from the observed LR image and regularized by the variation of directional gradients of the image patch. Experimental results demonstrate that our proposed method offers a remarkable quality improvement, both objectively and subjectively.",
keywords = "geometric similarity, Image interpolation, regularization, weighting coefficients",
author = "Shuyuan Zhu and Bing Zeng and Guanghui Liu and Liaoyuan Zeng and Lu Fang and Moncef Gabbouj",
year = "2015",
month = "8",
day = "4",
doi = "10.1109/ICME.2015.7177417",
language = "English",
isbn = "9781479970827",
booktitle = "2015 IEEE International Conference on Multimedia and Expo (ICME)",
publisher = "IEEE COMPUTER SOCIETY PRESS",

}

RIS (suitable for import to EndNote) - Download

TY - GEN

T1 - Image interpolation based on non-local geometric similarities

AU - Zhu, Shuyuan

AU - Zeng, Bing

AU - Liu, Guanghui

AU - Zeng, Liaoyuan

AU - Fang, Lu

AU - Gabbouj, Moncef

PY - 2015/8/4

Y1 - 2015/8/4

N2 - Image interpolation refers to constructing a high-resolution (HR) image from a low-resolution (LR) image. Traditionally, an HR image can be produced from an observed LR image via the polynomial-based interpolation (bi-linear or bi-cubic interpolations, involving a small number of neighbors around each interpolated position). The advanced interpolation makes use of the so-called 'geometric similarity' to design a set of optimal interpolation weighting coefficients. However, better geometric similarities can perhaps be found from a non-local area within the LR source image or even from other but similar images (possibly with higher resolutions). Based on this fact, we propose in this paper a non-local geometric similarity based interpolation scheme to construct HR images. In our proposed method, optimal weighting coefficients are determined by solving a regularized least squares problem which is built upon a number of dual reference patches drawn from the observed LR image and regularized by the variation of directional gradients of the image patch. Experimental results demonstrate that our proposed method offers a remarkable quality improvement, both objectively and subjectively.

AB - Image interpolation refers to constructing a high-resolution (HR) image from a low-resolution (LR) image. Traditionally, an HR image can be produced from an observed LR image via the polynomial-based interpolation (bi-linear or bi-cubic interpolations, involving a small number of neighbors around each interpolated position). The advanced interpolation makes use of the so-called 'geometric similarity' to design a set of optimal interpolation weighting coefficients. However, better geometric similarities can perhaps be found from a non-local area within the LR source image or even from other but similar images (possibly with higher resolutions). Based on this fact, we propose in this paper a non-local geometric similarity based interpolation scheme to construct HR images. In our proposed method, optimal weighting coefficients are determined by solving a regularized least squares problem which is built upon a number of dual reference patches drawn from the observed LR image and regularized by the variation of directional gradients of the image patch. Experimental results demonstrate that our proposed method offers a remarkable quality improvement, both objectively and subjectively.

KW - geometric similarity

KW - Image interpolation

KW - regularization

KW - weighting coefficients

UR - http://www.scopus.com/inward/record.url?scp=84946043968&partnerID=8YFLogxK

U2 - 10.1109/ICME.2015.7177417

DO - 10.1109/ICME.2015.7177417

M3 - Conference contribution

SN - 9781479970827

BT - 2015 IEEE International Conference on Multimedia and Expo (ICME)

PB - IEEE COMPUTER SOCIETY PRESS

ER -