A nuclear-norm model for multi-frame super-resolution reconstruction from video clips

Rui ZHAO; Raymond HF CHAN

doi:10.1007/978-3-030-04088-8_16

A nuclear-norm model for multi-frame super-resolution reconstruction from video clips

Rui ZHAO, Raymond HF CHAN^*

^*Corresponding author for this work

Research output: Book Chapters | Papers in Conference Proceedings › Book Chapter › Research › peer-review

Abstract

We propose a variational approach to obtain super-resolution images from multiple low-resolution frames extracted from video clips. First the displacement between the low-resolution frames and the reference frame is computed by an optical flow algorithm. Then a low-rank model is used to construct the reference frame in high resolution by incorporating the information of the low-resolution frames. The model has two terms: a 2-norm data fidelity term and a nuclear-norm regularization term. Alternating direction method of multipliers is used to solve the model. Comparison of our methods with other models on synthetic and real video clips shows that our resulting images are more accurate with less artifacts. It also provides much finer and discernable details.

Original language	English
Title of host publication	Structured Matrices in Numerical Linear Algebra: Analysis, Algorithms and Applications
Editors	Dario Andrea BINI, Fabio DI BENEDETTO, Eugene TYRTYSHNIKOV, Marc Van BAREL
Publisher	Springer, Cham
Pages	303-322
Number of pages	20
ISBN (Electronic)	9783030040888
ISBN (Print)	9783030040871
DOIs	https://doi.org/10.1007/978-3-030-04088-8_16
Publication status	Published - 2019
Externally published	Yes

Publication series

Name	Springer INdAM Series
Volume	30
ISSN (Print)	2281-518X
ISSN (Electronic)	2281-5198

Bibliographical note

Publisher Copyright:
© Springer Nature Switzerland AG 2019.

Keywords

Image processing
Low-rank approximation
Super-resolution

Access to Document

10.1007/978-3-030-04088-8_16

Cite this

ZHAO, R., & CHAN, R. HF. (2019). A nuclear-norm model for multi-frame super-resolution reconstruction from video clips. In D. A. BINI, F. DI. BENEDETTO, E. TYRTYSHNIKOV, & M. V. BAREL (Eds.), Structured Matrices in Numerical Linear Algebra: Analysis, Algorithms and Applications (pp. 303-322). (Springer INdAM Series; Vol. 30). Springer, Cham. https://doi.org/10.1007/978-3-030-04088-8_16

@inbook{276987b8e1ed47ef877e918ff3733bfe,

title = "A nuclear-norm model for multi-frame super-resolution reconstruction from video clips",

abstract = "We propose a variational approach to obtain super-resolution images from multiple low-resolution frames extracted from video clips. First the displacement between the low-resolution frames and the reference frame is computed by an optical flow algorithm. Then a low-rank model is used to construct the reference frame in high resolution by incorporating the information of the low-resolution frames. The model has two terms: a 2-norm data fidelity term and a nuclear-norm regularization term. Alternating direction method of multipliers is used to solve the model. Comparison of our methods with other models on synthetic and real video clips shows that our resulting images are more accurate with less artifacts. It also provides much finer and discernable details.",

keywords = "Image processing, Low-rank approximation, Super-resolution",

author = "Rui ZHAO and CHAN, {Raymond HF}",

note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2019.",

year = "2019",

doi = "10.1007/978-3-030-04088-8_16",

language = "English",

isbn = "9783030040871",

series = "Springer INdAM Series",

publisher = "Springer, Cham",

pages = "303--322",

editor = "BINI, {Dario Andrea} and BENEDETTO, {Fabio DI} and Eugene TYRTYSHNIKOV and BAREL, {Marc Van}",

booktitle = "Structured Matrices in Numerical Linear Algebra: Analysis, Algorithms and Applications",

}

A nuclear-norm model for multi-frame super-resolution reconstruction from video clips. / ZHAO, Rui; CHAN, Raymond HF.
Structured Matrices in Numerical Linear Algebra: Analysis, Algorithms and Applications. ed. / Dario Andrea BINI; Fabio DI BENEDETTO; Eugene TYRTYSHNIKOV; Marc Van BAREL. Springer, Cham, 2019. p. 303-322 (Springer INdAM Series; Vol. 30).

Research output: Book Chapters | Papers in Conference Proceedings › Book Chapter › Research › peer-review

TY - CHAP

T1 - A nuclear-norm model for multi-frame super-resolution reconstruction from video clips

AU - ZHAO, Rui

AU - CHAN, Raymond HF

N1 - Publisher Copyright: © Springer Nature Switzerland AG 2019.

PY - 2019

Y1 - 2019

N2 - We propose a variational approach to obtain super-resolution images from multiple low-resolution frames extracted from video clips. First the displacement between the low-resolution frames and the reference frame is computed by an optical flow algorithm. Then a low-rank model is used to construct the reference frame in high resolution by incorporating the information of the low-resolution frames. The model has two terms: a 2-norm data fidelity term and a nuclear-norm regularization term. Alternating direction method of multipliers is used to solve the model. Comparison of our methods with other models on synthetic and real video clips shows that our resulting images are more accurate with less artifacts. It also provides much finer and discernable details.

AB - We propose a variational approach to obtain super-resolution images from multiple low-resolution frames extracted from video clips. First the displacement between the low-resolution frames and the reference frame is computed by an optical flow algorithm. Then a low-rank model is used to construct the reference frame in high resolution by incorporating the information of the low-resolution frames. The model has two terms: a 2-norm data fidelity term and a nuclear-norm regularization term. Alternating direction method of multipliers is used to solve the model. Comparison of our methods with other models on synthetic and real video clips shows that our resulting images are more accurate with less artifacts. It also provides much finer and discernable details.

KW - Image processing

KW - Low-rank approximation

KW - Super-resolution

UR - http://www.scopus.com/inward/record.url?scp=85064740449&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-04088-8_16

DO - 10.1007/978-3-030-04088-8_16

M3 - Book Chapter

AN - SCOPUS:85064740449

SN - 9783030040871

T3 - Springer INdAM Series

SP - 303

EP - 322

BT - Structured Matrices in Numerical Linear Algebra: Analysis, Algorithms and Applications

A2 - BINI, Dario Andrea

A2 - BENEDETTO, Fabio DI

A2 - TYRTYSHNIKOV, Eugene

A2 - BAREL, Marc Van

PB - Springer, Cham

ER -

A nuclear-norm model for multi-frame super-resolution reconstruction from video clips

Abstract

Publication series

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this