Hyperspectral and color-infrared imaging from ultra-light aircraft: Potential to recognize tree species in urban environments

Gintautas Mozgeris (1) , Sébastien Gadal (2, 3) , Donatas Jonikavičius (1) , Lina Straigyte (1) , Walid Ouerghemmi (2, 4) , Vytautė Juodkienė (1)

1 Aleksandras Stulginskis University
2 ESPACE - Études des Structures, des Processus d’Adaptation et des Changements de l’Espace
3 AMU - Aix Marseille Université
4 CNRS - Centre National de la Recherche Scientifique

Imaging system based on simultaneous use of Rikola hyperspectral and RGB/NIR cameras installed on a manned ultra-light aircraft is introduced in this study. Simultaneously acquired hyperspectral and color- infrared (CIR) images were tested for their potential to identify deciduous tree species and estimate tree health in Kaunas city, Lithuania. Six urban deciduous tree species were separated using tree crown level statistics, extracted from 16 visible-near infrared spectral band hyperspectral images, and discriminant analyses with an overall classification accuracy of 63.1 %. Classification accuracy increased by 3 percent when hyperspectral images were integrated with simultaneously acquired CIR images. The accuracy in identifying tree health using fused hyperspectral and CIR images, ranged from poor to moderate.

Mots clés: hyperspectral imaging; color-infrared images; ultra-light aircraft; urban tree species identification; discriminant analysis

Domaines: Géographie; Environnement et Société; Méthodes et statistiques; Traitement des images [eess.IV]; Modélisation et simulation

Gintautas Mozgeris, Sébastien Gadal, Donatas Jonikavičius, Lina Straigyte, Walid Ouerghemmi, et al.. Hyperspectral and color-infrared imaging from ultra-light aircraft: Potential to recognize tree species in urban environments. 8th Workshop in Hyperspectral Image and Signal Processing: Evolution in Remote Sensing, Aug 2016, Los Angeles, United States. pp.542-546. ⟨hal-01359643⟩