Classification of Tree Species and Assessment of Oak Condition in a Mixed Broadleaf Forest Using Time Series of Hyperspatial Multispectral UAS Imagery

Abstract: Acute Oak Decline (AOD), a recently emerged disease in the UK, can cause tree mortality within a few years. Utilizing unmanned aerial systems (UAS) for the detection and monitoring the progression of such disease is highly attractive considering their potential for frequent and cost-effective deployment. To explore the possibility of sensing AOD induced stress with UAS acquired remote sensing data a woodland where symptoms of the disease were newly identified was chosen as an experimental site. The dense deciduous woodland is a natural reserve and features a rich species composition and high structural diversity. To investigate signals specific to oak species classification is hence undertaken. For this purpose, hyperspatial imagery (<10cm) of the woodland’s canopy was collected with a 5 band multispectral sensor.

Read the paper

Authors: Iglhaut, Jakob Swansea University

Associations: Swansea University