Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data

Paulo Silva Filho; Claudio Persello; Raian V. Maretto; Renato MacHado

doi:10.1109/IGARSS52108.2023.10282190

Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data

Paulo Silva Filho^*, Claudio Persello, Raian V. Maretto, Renato MacHado

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

2 Citations (Scopus)

57 Downloads (Pure)

Abstract

Despite its environmental and societal importance, accurately mapping the Brazilian Cerrado's vegetation is still an open challenge. Its diverse but spectrally similar physiognomies are difficult to be identified and mapped by state-of-the-art methods from only medium-to high-resolution optical images. This work investigates the fusion of Synthetic Aperture Radar (SAR) and optical data in convolutional neural network architectures to map the Cerrado according to a 2-level class hierarchy. Additionally, the proposed model is designed to deal with uncertainties that are brought by the difference in resolution between the input images (at 10m) and the reference data (at 30m). We tested four data fusion strategies and showed that the position for the data combination is important for the network to learn better features.

Original language	English
Title of host publication	IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Proceedings
Place of Publication	Piscataway, NJ
Publisher	IEEE
Pages	1365-1368
Number of pages	4
ISBN (Electronic)	979-8-3503-2010-7, 979-8-3503-2009-1 (USB)
ISBN (Print)	979-8-3503-3174-5
DOIs	https://doi.org/10.1109/IGARSS52108.2023.10282190
Publication status	Published - 2023
Event	43rd IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2023 - Pasadena Convention Center, Pasadena, United States Duration: 16 Jul 2023 → 21 Jul 2023 Conference number: 43 https://2023.ieeeigarss.org/index.php

Publication series

Name	International Geoscience and Remote Sensing Symposium (IGARSS)
Publisher	IEEE
Volume	2023
ISSN (Print)	2153-6996
ISSN (Electronic)	2153-7003

Conference

Conference	43rd IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2023
Abbreviated title	IGARSS 2023
Country/Territory	United States
City	Pasadena
Period	16/07/23 → 21/07/23
Internet address	https://2023.ieeeigarss.org/index.php

Keywords

Cerrado
Deep Learning (DL)
Remote sensing
SAR-optical data fusion
Semantic segmentation
2024 OA procedure

Access to Document

10.1109/IGARSS52108.2023.10282190

ArticleFinal published version, 1.7 MBLicence: Taverne

Cite this

Filho, P. S., Persello, C., Maretto, R. V., & MacHado, R. (2023). Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data. In IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Proceedings (pp. 1365-1368). (International Geoscience and Remote Sensing Symposium (IGARSS); Vol. 2023). IEEE. https://doi.org/10.1109/IGARSS52108.2023.10282190

Filho, Paulo Silva ; Persello, Claudio ; Maretto, Raian V. et al. / Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data. IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Proceedings. Piscataway, NJ : IEEE, 2023. pp. 1365-1368 (International Geoscience and Remote Sensing Symposium (IGARSS)).

@inproceedings{ee75d7c72d0d45009535f35570302ec9,

title = "Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data",

abstract = "Despite its environmental and societal importance, accurately mapping the Brazilian Cerrado's vegetation is still an open challenge. Its diverse but spectrally similar physiognomies are difficult to be identified and mapped by state-of-the-art methods from only medium-to high-resolution optical images. This work investigates the fusion of Synthetic Aperture Radar (SAR) and optical data in convolutional neural network architectures to map the Cerrado according to a 2-level class hierarchy. Additionally, the proposed model is designed to deal with uncertainties that are brought by the difference in resolution between the input images (at 10m) and the reference data (at 30m). We tested four data fusion strategies and showed that the position for the data combination is important for the network to learn better features.",

keywords = "Cerrado, Deep Learning (DL), Remote sensing, SAR-optical data fusion, Semantic segmentation, 2024 OA procedure",

author = "Filho, {Paulo Silva} and Claudio Persello and Maretto, {Raian V.} and Renato MacHado",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 43rd IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2023, IGARSS 2023 ; Conference date: 16-07-2023 Through 21-07-2023",

year = "2023",

doi = "10.1109/IGARSS52108.2023.10282190",

language = "English",

isbn = "979-8-3503-3174-5",

series = "International Geoscience and Remote Sensing Symposium (IGARSS)",

publisher = "IEEE",

pages = "1365--1368",

booktitle = "IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Proceedings",

address = "United States",

url = "https://2023.ieeeigarss.org/index.php",

}

Filho, PS, Persello, C , Maretto, RV & MacHado, R 2023, Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data. in IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Proceedings. International Geoscience and Remote Sensing Symposium (IGARSS), vol. 2023, IEEE, Piscataway, NJ, pp. 1365-1368, 43rd IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2023, Pasadena, California, United States, 16/07/23. https://doi.org/10.1109/IGARSS52108.2023.10282190

Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data. / Filho, Paulo Silva; Persello, Claudio ; Maretto, Raian V. et al.
IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Proceedings. Piscataway, NJ: IEEE, 2023. p. 1365-1368 (International Geoscience and Remote Sensing Symposium (IGARSS); Vol. 2023).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data

AU - Filho, Paulo Silva

AU - Persello, Claudio

AU - Maretto, Raian V.

AU - MacHado, Renato

N1 - Conference code: 43

PY - 2023

Y1 - 2023

N2 - Despite its environmental and societal importance, accurately mapping the Brazilian Cerrado's vegetation is still an open challenge. Its diverse but spectrally similar physiognomies are difficult to be identified and mapped by state-of-the-art methods from only medium-to high-resolution optical images. This work investigates the fusion of Synthetic Aperture Radar (SAR) and optical data in convolutional neural network architectures to map the Cerrado according to a 2-level class hierarchy. Additionally, the proposed model is designed to deal with uncertainties that are brought by the difference in resolution between the input images (at 10m) and the reference data (at 30m). We tested four data fusion strategies and showed that the position for the data combination is important for the network to learn better features.

AB - Despite its environmental and societal importance, accurately mapping the Brazilian Cerrado's vegetation is still an open challenge. Its diverse but spectrally similar physiognomies are difficult to be identified and mapped by state-of-the-art methods from only medium-to high-resolution optical images. This work investigates the fusion of Synthetic Aperture Radar (SAR) and optical data in convolutional neural network architectures to map the Cerrado according to a 2-level class hierarchy. Additionally, the proposed model is designed to deal with uncertainties that are brought by the difference in resolution between the input images (at 10m) and the reference data (at 30m). We tested four data fusion strategies and showed that the position for the data combination is important for the network to learn better features.

KW - Cerrado

KW - Deep Learning (DL)

KW - Remote sensing

KW - SAR-optical data fusion

KW - Semantic segmentation

KW - 2024 OA procedure

U2 - 10.1109/IGARSS52108.2023.10282190

DO - 10.1109/IGARSS52108.2023.10282190

M3 - Conference contribution

AN - SCOPUS:85178348700

SN - 979-8-3503-3174-5

T3 - International Geoscience and Remote Sensing Symposium (IGARSS)

SP - 1365

EP - 1368

BT - IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Proceedings

PB - IEEE

CY - Piscataway, NJ

T2 - 43rd IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2023

Y2 - 16 July 2023 through 21 July 2023

ER -

Filho PS, Persello C , Maretto RV, MacHado R. Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data. In IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Proceedings. Piscataway, NJ: IEEE. 2023. p. 1365-1368. (International Geoscience and Remote Sensing Symposium (IGARSS)). doi: 10.1109/IGARSS52108.2023.10282190

Investigating Sar-Optical Deep Learning Data Fusion to Map the Brazilian Cerrado Vegetation with Sentinel Data

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this