KIT | KIT-Bibliothek | Impressum | Datenschutz

Evaluation of Transformers and Convolutional Neural Networks for High-Dimensional Hyperspectral Soil Texture Classification

Kühnlein, Laura; Keller, Sina ORCID iD icon 1
1 Institut für Photogrammetrie und Fernerkundung (IPF), Karlsruher Institut für Technologie (KIT)

Abstract:

Soil texture is an important parameter influencing a multitude of ecosystem services. However, its determination in the laboratory is complex, time-consuming, and only reveals soil texture at a specific sampling point. Therefore, topsoil soil texture determined from space-borne remote sensing data offers advantages (areal and temporal availability, expanding possibilities with upcoming hyperspectral satellite systems).
Since no hyperspectral satellite data are available, we use hyperspectral reflectance data provided in the Land Use/Land Cover Area Frame Survey (LUCAS) dataset by the European Soil Data Centre. We resample the provided 4200 bands to the Environmental Mapping and Analysis Program (EnMAP) Resolution of 222 bands. Hereafter, we classify soil texture as sandy, silty, clayey, and loamy from these by applying distinct Transformer architecture as well as a one-dimensional convolutional neural network. Our best models multitemporal SimpleVIT and an ensemble approach score 65.89% and 67.62% overall accuracy, respectively.


Download
Originalveröffentlichung
DOI: 10.1109/WHISPERS56178.2022.9955087
Scopus
Zitationen: 1
Dimensions
Zitationen: 1
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Photogrammetrie und Fernerkundung (IPF)
Publikationstyp Proceedingsbeitrag
Publikationsjahr 2022
Sprache Englisch
Identifikator ISBN: 978-1-6654-7069-8
ISSN: 2158-6276
KITopen-ID: 1000150913
Erschienen in 2022 12th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS)
Veranstaltung 12th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS 2022), Rom, Italien, 13.09.2022 – 16.09.2022
Verlag Institute of Electrical and Electronics Engineers (IEEE)
Serie Workshop on Hyperspectral Image and Signal Processing, Evolution in Remote Sensing
Vorab online veröffentlicht am 22.11.2022
Nachgewiesen in Dimensions
Scopus
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page