Estimation of canopy attributes in beech forests using true colour digital images from a small fixed-wing UAV

Accurate estimates of forest canopy are essential for the characterization of forest ecosystems. Remotely-sensed techniques provide a unique way to obtain estimates over spatially extensive areas, but their application is limited by the spectral and temporal resolution available from these systems, which is often not suited to meet regional or local objectives. The use of unmanned aerial vehicles (UAV) as remote sensing platforms has recently gained increasing attention, but their applications in forestry are still at an experimental stage. In this study we described a methodology to obtain rapid and reliable estimates of forest canopy from a small UAV equipped with a commercial RGB camera. The red, green and blue digital numbers were converted to the green leaf algorithm (GLA) and to the CIE L*a*b*colour space to obtain estimates of canopy cover, foliage clumping and leaf area index (L) from aerial images. Canopy attributes were compared with in situ estimates obtained from two digital canopy photographic techniques (cover and fisheye photography). The method was tested in beech forests. UAV images accurately quantified canopy cover even in very dense stand conditions, despite a tendency to not detecting small within-crown gaps in aerial images, leading to a measurement of a quantity much closer to crown cover estimated from in situ cover photography. Estimates of L from UAV images significantly agreed with that obtained from fisheye images, but the accuracy of UAV estimates is influenced by the appropriate assumption of leaf angle distribution. We concluded that true colour UAV images can be effectively used to obtain rapid, cheap and meaningful estimates of forest canopy attributes at medium-large scales. UAV can combine the advantage of high resolution imagery with quick turnaround series, being therefore suitable for routine forest stand monitoring and real-time applications