A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery

There is a growing interest in the development of automated data processing workflows that provide reliable, high spatial resolution land cover maps. However, high-resolution remote sensing images are not always affordable. Taking into account the free availability of Sentinel-2 satellite data, in t...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Saüc Abadal, Luis Salgueiro, Javier Marcello, Verónica Vilaplana
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	super-resolution semantic segmentation deep learning convolutional neural network Sentinel-2 Science Q
Acceso en línea:	https://doaj.org/article/b38e9069a39e4b2b9a813c90ee6d5ccc
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:b38e9069a39e4b2b9a813c90ee6d5ccc
record_format	dspace
spelling	oai:doaj.org-article:b38e9069a39e4b2b9a813c90ee6d5ccc2021-11-25T18:54:12ZA Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery10.3390/rs132245472072-4292https://doaj.org/article/b38e9069a39e4b2b9a813c90ee6d5ccc2021-11-01T00:00:00Zhttps://www.mdpi.com/2072-4292/13/22/4547https://doaj.org/toc/2072-4292There is a growing interest in the development of automated data processing workflows that provide reliable, high spatial resolution land cover maps. However, high-resolution remote sensing images are not always affordable. Taking into account the free availability of Sentinel-2 satellite data, in this work we propose a deep learning model to generate high-resolution segmentation maps from low-resolution inputs in a multi-task approach. Our proposal is a dual-network model with two branches: the Single Image Super-Resolution branch, that reconstructs a high-resolution version of the input image, and the Semantic Segmentation Super-Resolution branch, that predicts a high-resolution segmentation map with a scaling factor of 2. We performed several experiments to find the best architecture, training and testing on a subset of the S2GLC 2017 dataset. We based our model on the DeepLabV3+ architecture, enhancing the model and achieving an improvement of 5% on IoU and almost 10% on the recall score. Furthermore, our qualitative results demonstrate the effectiveness and usefulness of the proposed approach.Saüc AbadalLuis SalgueiroJavier MarcelloVerónica VilaplanaMDPI AGarticlesuper-resolutionsemantic segmentationdeep learningconvolutional neural networkSentinel-2ScienceQENRemote Sensing, Vol 13, Iss 4547, p 4547 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	super-resolution semantic segmentation deep learning convolutional neural network Sentinel-2 Science Q
spellingShingle	super-resolution semantic segmentation deep learning convolutional neural network Sentinel-2 Science Q Saüc Abadal Luis Salgueiro Javier Marcello Verónica Vilaplana A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery
description	There is a growing interest in the development of automated data processing workflows that provide reliable, high spatial resolution land cover maps. However, high-resolution remote sensing images are not always affordable. Taking into account the free availability of Sentinel-2 satellite data, in this work we propose a deep learning model to generate high-resolution segmentation maps from low-resolution inputs in a multi-task approach. Our proposal is a dual-network model with two branches: the Single Image Super-Resolution branch, that reconstructs a high-resolution version of the input image, and the Semantic Segmentation Super-Resolution branch, that predicts a high-resolution segmentation map with a scaling factor of 2. We performed several experiments to find the best architecture, training and testing on a subset of the S2GLC 2017 dataset. We based our model on the DeepLabV3+ architecture, enhancing the model and achieving an improvement of 5% on IoU and almost 10% on the recall score. Furthermore, our qualitative results demonstrate the effectiveness and usefulness of the proposed approach.
format	article
author	Saüc Abadal Luis Salgueiro Javier Marcello Verónica Vilaplana
author_facet	Saüc Abadal Luis Salgueiro Javier Marcello Verónica Vilaplana
author_sort	Saüc Abadal
title	A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery
title_short	A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery
title_full	A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery
title_fullStr	A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery
title_full_unstemmed	A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery
title_sort	dual network for super-resolution and semantic segmentation of sentinel-2 imagery
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/b38e9069a39e4b2b9a813c90ee6d5ccc
work_keys_str_mv	AT saucabadal adualnetworkforsuperresolutionandsemanticsegmentationofsentinel2imagery AT luissalgueiro adualnetworkforsuperresolutionandsemanticsegmentationofsentinel2imagery AT javiermarcello adualnetworkforsuperresolutionandsemanticsegmentationofsentinel2imagery AT veronicavilaplana adualnetworkforsuperresolutionandsemanticsegmentationofsentinel2imagery AT saucabadal dualnetworkforsuperresolutionandsemanticsegmentationofsentinel2imagery AT luissalgueiro dualnetworkforsuperresolutionandsemanticsegmentationofsentinel2imagery AT javiermarcello dualnetworkforsuperresolutionandsemanticsegmentationofsentinel2imagery AT veronicavilaplana dualnetworkforsuperresolutionandsemanticsegmentationofsentinel2imagery
_version_	1718410613713010688

A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery

Ejemplares similares