Light U-Net with a New Morphological Attention Gate Model Application to Analyse Wood Sections

Rémi Decelle, Phuc Ngo, Isabelle Debled-Rennesson, Frédéric Mothe, Fleur Longuetaud

2023

Abstract

This article focuses on heartwood segmentation from cross-section RGB images (see Fig.1). In this context, we propose a novel attention gate (AG) model for both improving performance and making light convolutional neural networks (CNNs). Our proposed AG is based on mathematical morphology operators. Our light CNN is based on the U-Net architecture and called Light U-net (LU-Net). Experimental results show that AGs consistently improve the prediction performance of LU-Net across different wood cross-section datasets. Our proposed morphological AG achieves better performance than original U-Net with 10 times less parameters.

Paper Citation

in Harvard Style

Decelle R., Ngo P., Debled-Rennesson I., Mothe F. and Longuetaud F. (2023). Light U-Net with a New Morphological Attention Gate Model Application to Analyse Wood Sections. In Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-626-2, pages 759-766. DOI: 10.5220/0011626800003411

in Bibtex Style

@conference{icpram23,
author={Rémi Decelle and Phuc Ngo and Isabelle Debled-Rennesson and Frédéric Mothe and Fleur Longuetaud},
title={Light U-Net with a New Morphological Attention Gate Model Application to Analyse Wood Sections},
booktitle={Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2023},
pages={759-766},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011626800003411},
isbn={978-989-758-626-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - Light U-Net with a New Morphological Attention Gate Model Application to Analyse Wood Sections
SN - 978-989-758-626-2
AU - Decelle R.
AU - Ngo P.
AU - Debled-Rennesson I.
AU - Mothe F.
AU - Longuetaud F.
PY - 2023
SP - 759
EP - 766
DO - 10.5220/0011626800003411