Abstract:
An old-growth plantation is an iconic site that has great value. Assessing stand
parameters is laborious and time-consuming in the field. Therefore, this case study
aimed to estimate the stand parameters of a 114-year-old Japanese larch (Larix
kaempheri) using UAV photogrammetry in Northern Japan. First, 3D point clouds and
orthomosaic were generated through SfM technology from the imagery at two UAV
altitudes, 80 m, and 120 m. Then, a canopy height model (CHM) was generated by
subtracting LiDAR DTM from UAV DSM. We used a Shiny-based R package for
individual tree detection (ITD), tree height (TH) and crown area (CA) from the CHM. We
also used local maxima algorithm for ITD. Stand density, mean TH, and CA cover were
estimated from ITD, TH and CA. Results of accuracy were tested using F-Score, R 2 and
RMSE against field data and manually delineated CA. The maximum height of CHM
was 43.66 m and 44.21 m at the UAV flight 80 m and 120 m, respectively whereas the
field maximum TH was 42.9 m.
