Dual-grid parameter choice method with application to image deblurring

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Juvonen, Markus, Jensen, Bjørn, Pohjola, Ilmari, Dong, Yiqiu, Siltanen, Samuli
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917072244572160
author Juvonen, Markus
Jensen, Bjørn
Pohjola, Ilmari
Dong, Yiqiu
Siltanen, Samuli
author_facet Juvonen, Markus
Jensen, Bjørn
Pohjola, Ilmari
Dong, Yiqiu
Siltanen, Samuli
contents Variational regularization of ill-posed inverse problems is based on minimizing the sum of a data fidelity term and a regularization term. The balance between them is tuned using a positive regularization parameter, whose automatic choice remains an open question in general. A novel approach for parameter choice is introduced, based on the use of two slightly different computational models for the same inverse problem. Small parameter values should give two very different reconstructions due to amplification of noise. Large parameter values lead to two identical but trivial reconstructions. Optimal parameter is chosen between the extremes by matching image similarity of the two reconstructions with a pre-defined value. Efficacy of the new method is demonstrated with image deblurring using measured data and two different regularizers.
format Preprint
id arxiv_https___arxiv_org_abs_2504_10259
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Dual-grid parameter choice method with application to image deblurring
Juvonen, Markus
Jensen, Bjørn
Pohjola, Ilmari
Dong, Yiqiu
Siltanen, Samuli
Numerical Analysis
65-XX, 00A69
Variational regularization of ill-posed inverse problems is based on minimizing the sum of a data fidelity term and a regularization term. The balance between them is tuned using a positive regularization parameter, whose automatic choice remains an open question in general. A novel approach for parameter choice is introduced, based on the use of two slightly different computational models for the same inverse problem. Small parameter values should give two very different reconstructions due to amplification of noise. Large parameter values lead to two identical but trivial reconstructions. Optimal parameter is chosen between the extremes by matching image similarity of the two reconstructions with a pre-defined value. Efficacy of the new method is demonstrated with image deblurring using measured data and two different regularizers.
title Dual-grid parameter choice method with application to image deblurring
topic Numerical Analysis
65-XX, 00A69
url https://arxiv.org/abs/2504.10259