Trainable Regularization for Multi-frame Superresolution

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

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandBegutachtung

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.
Originalspracheenglisch
TitelPattern Recognition
UntertitelGerman Conference, GCPR 2017, Proceedings
Redakteure/-innenV. Roth, T. Vetter
Herausgeber (Verlag)Springer
Seiten90-100
ISBN (Print)978-3-319-66708-9
DOIs
PublikationsstatusVeröffentlicht - 2017

Publikationsreihe

NameLecture Notes in Computer Science
Band10496

Fingerprint

Untersuchen Sie die Forschungsthemen von „Trainable Regularization for Multi-frame Superresolution“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren