KIT | KIT-Bibliothek | Impressum | Datenschutz

Soil Texture Classification with 1D Convolutional Neural Networks based on Hyperspectral Data

Riese, Felix M.; Keller, Sina

Abstract:
Soil texture is important for many environmental processes. In this paper, we study the classification of soil texture based on hyperspectral data. We develop and implement three 1-dimensional (1D) convolutional neural networks (CNN): the LucasCNN, the LucasResNet which contains an identity block as residual network, and the LucasCoordConv with an additional coordinates layer. Furthermore, we modify two existing 1D CNN approaches for the presented classification task. The code of all five CNN approaches is available on GitHub (Riese, 2019). We evaluate the performance of the CNN approaches and compare them to a random forest classifier. Thereby, we rely on the freely available LUCAS topsoil dataset. The CNN approach with the least depth turns out to be the best performing classifier. The LucasCoordConv achieves the best performance regarding the average accuracy. In future work, we can further enhance the introduced LucasCNN, LucasResNet and LucasCoordConv and include additional variables of the rich LUCAS dataset.

Open Access Logo


Verlagsausgabe §
DOI: 10.5445/IR/1000089586
Veröffentlicht am 07.06.2019
Originalveröffentlichung
DOI: 10.5194/isprs-annals-IV-2-W5-615-2019
Coverbild
Zugehörige Institution(en) am KIT Institut für Photogrammetrie und Fernerkundung (IPF)
Publikationstyp Proceedingsbeitrag
Jahr 2019
Sprache Englisch
Identifikator KITopen-ID: 1000089586
Erschienen in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences (2019 ISPRS) Geospatial Week 2019, Enschede, NL, June 10-14, 2019. Vol. IV-2/W5
Seiten 615-621
Vorab online veröffentlicht am 15.01.2019
Schlagworte Soil Texture, Hyperspectral, Machine Learning, CNN Residual Network, CoordConv
Nachgewiesen in Scopus
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page