A Deep Learning Architecture for Land Cover Mapping Using Spatio-Temporal Sentinel-1 Features

Land cover (LC) mapping using satellite imagery is critical for environmental monitoring and management. Deep learning (DL), particularly convolutional neural networks (CNNs) and vision transformers (ViTs), have revolutionized this field by enhancing the accuracy of classification tasks. In this wor...

Full description

Saved in:
Bibliographic Details
Main Authors: Luigi Russo, Antonietta Sorriso, Silvia Liberata Ullo, Paolo Gamba
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10948270/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850173833682616320
author Luigi Russo
Antonietta Sorriso
Silvia Liberata Ullo
Paolo Gamba
author_facet Luigi Russo
Antonietta Sorriso
Silvia Liberata Ullo
Paolo Gamba
author_sort Luigi Russo
collection DOAJ
description Land cover (LC) mapping using satellite imagery is critical for environmental monitoring and management. Deep learning (DL), particularly convolutional neural networks (CNNs) and vision transformers (ViTs), have revolutionized this field by enhancing the accuracy of classification tasks. In this work, a novel approach combining a transformer-based Swin-Unet architecture with seasonal synthesized spatio-temporal images has been employed to classify LC types using spatio-temporal features extracted from Sentinel-1 (S1) synthetic aperture radar (SAR) data, organized into seasonal clusters. The study focuses on three distinct regions—Amazonia, Africa, and Siberia—and evaluates the model performance across diverse ecoregions within these areas. By utilizing seasonal feature sequences instead of dense temporal sequences, notable performance improvements have been achieved, especially in regions with temporal data gaps such as Siberia, where S1 data distribution is uneven and nonuniform. The results demonstrate the effectiveness and the generalization capabilities of the proposed methodology in achieving high overall accuracy (OA) values, even in regions with limited training data.
format Article
id doaj-art-3efae2d148b24d12b30502bb7b687fda
institution OA Journals
issn 1939-1404
2151-1535
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-3efae2d148b24d12b30502bb7b687fda2025-08-20T02:19:47ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-0118105621058110.1109/JSTARS.2025.355768710948270A Deep Learning Architecture for Land Cover Mapping Using Spatio-Temporal Sentinel-1 FeaturesLuigi Russo0https://orcid.org/0009-0001-3659-0087Antonietta Sorriso1https://orcid.org/0009-0005-9892-2445Silvia Liberata Ullo2https://orcid.org/0000-0001-6294-0581Paolo Gamba3https://orcid.org/0000-0002-9576-6337Department of Electrical, Computer and Biomedical Engineering, University of Pavia, Pavia, ItalyDepartment of Electrical, Computer and Biomedical Engineering, University of Pavia, Pavia, ItalyDepartment of Engineering, University of Sannio, Benevento, ItalyDepartment of Electrical, Computer and Biomedical Engineering, University of Pavia, Pavia, ItalyLand cover (LC) mapping using satellite imagery is critical for environmental monitoring and management. Deep learning (DL), particularly convolutional neural networks (CNNs) and vision transformers (ViTs), have revolutionized this field by enhancing the accuracy of classification tasks. In this work, a novel approach combining a transformer-based Swin-Unet architecture with seasonal synthesized spatio-temporal images has been employed to classify LC types using spatio-temporal features extracted from Sentinel-1 (S1) synthetic aperture radar (SAR) data, organized into seasonal clusters. The study focuses on three distinct regions—Amazonia, Africa, and Siberia—and evaluates the model performance across diverse ecoregions within these areas. By utilizing seasonal feature sequences instead of dense temporal sequences, notable performance improvements have been achieved, especially in regions with temporal data gaps such as Siberia, where S1 data distribution is uneven and nonuniform. The results demonstrate the effectiveness and the generalization capabilities of the proposed methodology in achieving high overall accuracy (OA) values, even in regions with limited training data.https://ieeexplore.ieee.org/document/10948270/Convolutional neural network (CNN)deep learning (DL)land cover (LC) mappingneural networkSentinel-1 (S1)swin-Unet
spellingShingle Luigi Russo
Antonietta Sorriso
Silvia Liberata Ullo
Paolo Gamba
A Deep Learning Architecture for Land Cover Mapping Using Spatio-Temporal Sentinel-1 Features
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Convolutional neural network (CNN)
deep learning (DL)
land cover (LC) mapping
neural network
Sentinel-1 (S1)
swin-Unet
title A Deep Learning Architecture for Land Cover Mapping Using Spatio-Temporal Sentinel-1 Features
title_full A Deep Learning Architecture for Land Cover Mapping Using Spatio-Temporal Sentinel-1 Features
title_fullStr A Deep Learning Architecture for Land Cover Mapping Using Spatio-Temporal Sentinel-1 Features
title_full_unstemmed A Deep Learning Architecture for Land Cover Mapping Using Spatio-Temporal Sentinel-1 Features
title_short A Deep Learning Architecture for Land Cover Mapping Using Spatio-Temporal Sentinel-1 Features
title_sort deep learning architecture for land cover mapping using spatio temporal sentinel 1 features
topic Convolutional neural network (CNN)
deep learning (DL)
land cover (LC) mapping
neural network
Sentinel-1 (S1)
swin-Unet
url https://ieeexplore.ieee.org/document/10948270/
work_keys_str_mv AT luigirusso adeeplearningarchitectureforlandcovermappingusingspatiotemporalsentinel1features
AT antoniettasorriso adeeplearningarchitectureforlandcovermappingusingspatiotemporalsentinel1features
AT silvialiberataullo adeeplearningarchitectureforlandcovermappingusingspatiotemporalsentinel1features
AT paologamba adeeplearningarchitectureforlandcovermappingusingspatiotemporalsentinel1features
AT luigirusso deeplearningarchitectureforlandcovermappingusingspatiotemporalsentinel1features
AT antoniettasorriso deeplearningarchitectureforlandcovermappingusingspatiotemporalsentinel1features
AT silvialiberataullo deeplearningarchitectureforlandcovermappingusingspatiotemporalsentinel1features
AT paologamba deeplearningarchitectureforlandcovermappingusingspatiotemporalsentinel1features