Trainable Regularization for Multi-frame Superresolution

Teresa Klatzer; Daniel Soukup; Erich Kobler; Kerstin Hammernik; Thomas Pock

doi:10.1007/978-3-319-66709-6_8

Trainable Regularization for Multi-frame Superresolution

Teresa Klatzer, Daniel Soukup, Erich Kobler, Kerstin Hammernik, Thomas Pock

Institute of Computer Graphics and Vision (7100)

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Abstract

In this paper, we present a novel method for multi-frame superresolution (SR). Our main goal is to improve the spatial resolution of a multi-line scan camera for an industrial inspection task. High resolution output images are reconstructed using our proposed SR algorithm for multi-channel data, which is based on the trainable reaction-diffusion model. As this is a supervised learning approach, we simulate ground truth data for a real imaging scenario. We show that learning a regularizer for the SR problem improves the reconstruction results compared to
an iterative reconstruction algorithm using TV or TGV regularization. We test the learned regularizer, trained on simulated data, on images acquired with the real camera setup and achieve excellent results.

Original language	English
Title of host publication	Pattern Recognition
Subtitle of host publication	German Conference, GCPR 2017, Proceedings
Editors	V. Roth, T. Vetter
Publisher	Springer
Pages	90-100
ISBN (Print)	978-3-319-66708-9
DOIs	https://doi.org/10.1007/978-3-319-66709-6_8
Publication status	Published - 2017

Publication series

Name	Lecture Notes in Computer Science
Volume	10496

Access to Document

10.1007/978-3-319-66709-6_8

learnedmultiframesr

Cite this

Trainable Regularization for Multi-frame Superresolution. / Klatzer, Teresa; Soukup, Daniel; Kobler, Erich et al.
Pattern Recognition: German Conference, GCPR 2017, Proceedings. ed. / V. Roth; T. Vetter. Springer, 2017. p. 90-100 (Lecture Notes in Computer Science; Vol. 10496).

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

@inproceedings{b01ca01cda754b9683648dac43549b62,

title = "Trainable Regularization for Multi-frame Superresolution",

abstract = "In this paper, we present a novel method for multi-frame superresolution (SR). Our main goal is to improve the spatial resolution of a multi-line scan camera for an industrial inspection task. High resolution output images are reconstructed using our proposed SR algorithm for multi-channel data, which is based on the trainable reaction-diffusion model. As this is a supervised learning approach, we simulate ground truth data for a real imaging scenario. We show that learning a regularizer for the SR problem improves the reconstruction results compared toan iterative reconstruction algorithm using TV or TGV regularization. We test the learned regularizer, trained on simulated data, on images acquired with the real camera setup and achieve excellent results.",

author = "Teresa Klatzer and Daniel Soukup and Erich Kobler and Kerstin Hammernik and Thomas Pock",

year = "2017",

doi = "10.1007/978-3-319-66709-6_8",

language = "English",

isbn = "978-3-319-66708-9",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "90--100",

editor = "V. Roth and T. Vetter",

booktitle = "Pattern Recognition",

}

TY - GEN

T1 - Trainable Regularization for Multi-frame Superresolution

AU - Klatzer, Teresa

AU - Soukup, Daniel

AU - Kobler, Erich

AU - Hammernik, Kerstin

AU - Pock, Thomas

PY - 2017

Y1 - 2017

N2 - In this paper, we present a novel method for multi-frame superresolution (SR). Our main goal is to improve the spatial resolution of a multi-line scan camera for an industrial inspection task. High resolution output images are reconstructed using our proposed SR algorithm for multi-channel data, which is based on the trainable reaction-diffusion model. As this is a supervised learning approach, we simulate ground truth data for a real imaging scenario. We show that learning a regularizer for the SR problem improves the reconstruction results compared toan iterative reconstruction algorithm using TV or TGV regularization. We test the learned regularizer, trained on simulated data, on images acquired with the real camera setup and achieve excellent results.

AB - In this paper, we present a novel method for multi-frame superresolution (SR). Our main goal is to improve the spatial resolution of a multi-line scan camera for an industrial inspection task. High resolution output images are reconstructed using our proposed SR algorithm for multi-channel data, which is based on the trainable reaction-diffusion model. As this is a supervised learning approach, we simulate ground truth data for a real imaging scenario. We show that learning a regularizer for the SR problem improves the reconstruction results compared toan iterative reconstruction algorithm using TV or TGV regularization. We test the learned regularizer, trained on simulated data, on images acquired with the real camera setup and achieve excellent results.

U2 - 10.1007/978-3-319-66709-6_8

DO - 10.1007/978-3-319-66709-6_8

M3 - Conference paper

SN - 978-3-319-66708-9

T3 - Lecture Notes in Computer Science

SP - 90

EP - 100

BT - Pattern Recognition

A2 - Roth, V.

A2 - Vetter, T.

PB - Springer

ER -

Trainable Regularization for Multi-frame Superresolution

Abstract

Publication series

Access to Document

Fingerprint

Cite this