Per-channel regularization for regression-based spectral reconstruction

Lin, Yi Tun and Finlayson, Graham D. (2020) Per-channel regularization for regression-based spectral reconstruction. CEUR Workshop Proceedings, 2688. ISSN 1613-0073

[img]
Preview
PDF (Published_Version) - Published Version
Available under License Creative Commons Attribution.

Download (20MB) | Preview

Abstract

Spectral reconstruction algorithms seek to recover spectra from RGB images. This estimation problem is often formulated as least-squares regression, and a Tikhonov regularization is generally incorporated, both to support stable estimation in the presence of noise and to prevent over-fitting. The degree of regularization is controlled by a single penalty-term parameter, which is often selected using the cross validation experimental methodology. In this paper, we generalize the simple regularization approach to admit a per-spectral-channel optimization setting, and a modified cross-validation procedure is developed. Experiments validate our method. Compared to the conventional regularization, our per-channel approach significantly improves the reconstruction accuracy at multiple spectral channels, by up to 17% increments for all the considered models.

Item Type: Article
Uncontrolled Keywords: hyperspectral imaging,multispectral imaging,spectral reconstruction,computer science(all) ,/dk/atira/pure/subjectarea/asjc/1700
Faculty \ School: Faculty of Science > School of Computing Sciences
Related URLs:
Depositing User: LivePure Connector
Date Deposited: 14 Nov 2020 01:15
Last Modified: 16 Jan 2021 00:53
URI: https://ueaeprints.uea.ac.uk/id/eprint/77689
DOI:

Actions (login required)

View Item View Item