Can iPhone/iPad LiDAR data improve canopy height model derived from UAV?

Abstract

Aerial images resulting from unmanned aerial vehicle (UAV) are widely used to estimate tree height. The filtering method is required to distinguish between ground and off-ground point clouds to generate a canopy height model. However, the filtering method is not always perfect since UAV data cannot penetrate canopies into the forest floor. The release of iPhone/iPad devices with built-in LiDAR sensors enables the more affordable use of LiDAR for forestry study, including the measurement of local topography below forest stands. This study investigates to what extent iPhone/iPad LiDAR can improve the accuracy of canopy height model from the UAV. The integration of UAV and iPhone/iPad LiDAR data managed to increase the accuracy of tree height model with a mean absolute error (MAE) of 2.188 m, compared to UAV data (MAE = 2.446 m). This preliminary study showed the potential of combining UAV and iPhone/iPad LiDAR data for estimating tree height.