Efficient and Accurate Hyperspectral Image Demosaicing with Neural Network Architectures

Eric Wisotzky, Eric Wisotzky, Lara Wallburg, Anna Hilsmann, Peter Eisert, Peter Eisert, Thomas Wittenberg, Thomas Wittenberg, Stephan Göb, Stephan Göb

2024

Abstract

Neural network architectures for image demosaicing have been become more and more complex. This results in long training periods of such deep networks and the size of the networks is huge. These two factors prevent practical implementation and usage of the networks in real-time platforms, which generally only have limited resources. This study investigates the effectiveness of neural network architectures in hyperspectral image demosaicing. We introduce a range of network models and modifications, and compare them with classical interpolation methods and existing reference network approaches. The aim is to identify robust and efficient performing network architectures. Our evaluation is conducted on two datasets, ”SimpleData” and ”SimReal-Data,” representing different degrees of realism in multispectral filter array (MSFA) data. The results indicate that our networks outperform or match reference models in both datasets demonstrating exceptional performance. Notably, our approach focuses on achieving correct spectral reconstruction rather than just visual appeal, and this emphasis is supported by quantitative and qualitative assessments. Furthermore, our findings suggest that efficient demosaicing solutions, which require fewer parameters, are essential for practical applications. This research contributes valuable insights into hyperspectral imaging and its potential applications in various fields, including medical imaging.

Paper Citation

in Harvard Style

Wisotzky E., Wallburg L., Hilsmann A., Eisert P., Wittenberg T. and Göb S. (2024). Efficient and Accurate Hyperspectral Image Demosaicing with Neural Network Architectures. In Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 3: VISAPP; ISBN 978-989-758-679-8, SciTePress, pages 541-550. DOI: 10.5220/0012392300003660

in Bibtex Style

@conference{visapp24,
author={Eric Wisotzky and Lara Wallburg and Anna Hilsmann and Peter Eisert and Thomas Wittenberg and Stephan Göb},
title={Efficient and Accurate Hyperspectral Image Demosaicing with Neural Network Architectures},
booktitle={Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 3: VISAPP},
year={2024},
pages={541-550},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012392300003660},
isbn={978-989-758-679-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 3: VISAPP
TI - Efficient and Accurate Hyperspectral Image Demosaicing with Neural Network Architectures
SN - 978-989-758-679-8
AU - Wisotzky E.
AU - Wallburg L.
AU - Hilsmann A.
AU - Eisert P.
AU - Wittenberg T.
AU - Göb S.
PY - 2024
SP - 541
EP - 550
DO - 10.5220/0012392300003660
PB - SciTePress