Détail de l'auteur
Auteur Patrick Launeau |
Documents disponibles écrits par cet auteur (1)
trié(s) par (Pertinence décroissant(e), Titre croissant(e)) Affiner la recherche
Contribution of Sentinel-2 satellite images for habitat mapping of the Natura 2000 site "Estuaire de la Loire" (France) / Mathieu Le Dez in Remote Sensing Applications : Society and Environment, vol. 24 (Année 2021)
[article]
Titre : Contribution of Sentinel-2 satellite images for habitat mapping of the Natura 2000 site "Estuaire de la Loire" (France) Type de document : texte imprimé Auteurs : Mathieu Le Dez, Auteur ; Marc Robin, Auteur ; Patrick Launeau, Auteur Année de publication : 2021 Article en page(s) : 100637 (15 p.) Langues : Anglais (eng) Catégories : [habitats/milieux] 13 - Estuaires et rivières tidales (soumises à marées)
[Thèmes] Cartographie des habitats
[Thèmes] Images satellites
[Thèmes] Natura 2000
[Thèmes] TélédétectionMots-clés : Sentinel-2 EUNIS phytosociologie Résumé : "Habitat mapping is an essential tool to implement the European Habitats Directive 92/43/EEC and to manage the Natura 2000 protected areas network. However, their elaboration is most often based on field surveys, which are time-consuming and expensive when dealing with large areas. In this study, we aimed to evaluate the contribution of Sentinel-2 satellite images for mapping habitats of the Natura 2000 site ‘Estuaire de la Loire’. We used 1,248 phytosociological plots to establish a complete typology of the study site's habitats. Locating these plots led us to select training areas that characterize 39 habitats to calibrate the Random Forest algorithm. We classified 11 single-date images spread evenly over a year and performed a multi-date classification of 9 images. We obtained high overall accuracy between 76.47% and 87.28% for the single-date image classifications and 98.7% for multi-date image classification. Our results demonstrate that Sentinel-2 images are appropriate for accurate habitat mapping and constitute a relevant tool to identify and conserve habitats of community interest." (source : auteurs) Note de contenu : "Highlights :
- We classified habitats of a French Natura 2000 site using Sentinel-2 imagery.
- Single-date image classification yielded up to 87.28% overall accuracy.
- Use of multi-date imagery classification increased this accuracy up to 98.7%.
- Sentinel-2 images are not suitable for mapping habitats that occupy small areas."Type de publication : périodique Référence biblio : Le Dez M., Robin M., Launeau P., 2021 - Contribution of Sentinel-2 satellite images for habitat mapping of the Natura 2000 site "Estuaire de la Loire" (France). Remote Sensing Applications : Society and Environment, 24 : 100637 (15 p.). ID PMB : 70129 DOI : 10.1016/j.rsase.2021.100637 Permalink : http://www.cbnbrest.fr/catalogue_en_ligne/index.php?lvl=notice_display&id=70129
in Remote Sensing Applications : Society and Environment > vol. 24 (Année 2021) . - 100637 (15 p.)[article]Exemplaires(0)
Disponibilité aucun exemplaire