Improving the classification of urban tree diversity from Very High Spatial Resolution hyperspectral images: comparison of multiples techniques

Charlotte Brabant (1) , Emilien Alvarez-Vanhard (1) , Thomas Houet (1)

1 LETG - Rennes - Littoral, Environnement, Télédétection, Géomatique

The aim of the study is to compare and assess the efficiency of conventional hyperspectral techniques (dimension reduction, learning and classification methods) to classify the urban tree vegetation diversity. A specific focus is made using very high spatial resolution hyperspectral images acquired from airborne sensor and simulated at various spatial/spectral resolutions. Results show that 78.8% of overall accuracy can be reached using MNF and SVM methods to classify urban tree diversity. Moreover, results are more sensitive to learning methods rather than dimension reduction or classification ones. Finally, VHRS hyperspectral imagery are promising for such a key urban environmental issue.

Mots clés: classification; reduction dimension; learning methods; Tree species; urban area

Domaines: Etudes de l'environnement; Traitement du signal et de l'image [eess.SP]

Charlotte Brabant, Emilien Alvarez-Vanhard, Thomas Houet. Improving the classification of urban tree diversity from Very High Spatial Resolution hyperspectral images: comparison of multiples techniques. Joint Urban Remote Sensing Event (JURSE 2019), May 2019, Vannes, France. ⟨halshs-02191097⟩