Lamprecht, Sebastian (2015)

aTrunk - An ALS-based Trunk Detection Algorithm

This paper presents a novel tree trunk detection approach for high resolution multi pulse airborne LiDAR (Light Detection And Ranging).  The multi-core Divide & Conquer algorithm uses a 3D-clustering approach to isolate points associated with single trunks directly out of the raw point cloud. For each trunk, a principal-component-based linear model is fitted, while a modification of LO-RANSAC is used to identify an optimal model. The algorithm returns  a  vector based model for each identified trunk while parameters like the ground position, zenith orientation, azimuth orientation and length of the trunk are additionally provided. The algorithm performed well for a study area of 109 trees (about 2/3 spruce and 1/3 beech), with  a point density of around 7.6 points per m², while a detection rate of about 70% with an average difference in positioning of 0.32m and an RMSE of 0.42m is reached.