Identification of individual tree crowns from LiDAR data using a circle fitting algorithm with local maxima and minima filtering

Abstract

In this study, we propose an algorithm for identifying tree crowns from LiDAR data based on the geometric relationship between local maxima and minima in forests. The local maxima and minima of LiDAR data were extracted as tree tops and crown boundaries, respectively. The most reasonable circles estimated from four local minima closest to the tree top were fitted as tree crowns. We identified 77% of the reference tree crowns using LiDAR data from dense and mixed forests in Korea, with a point density of approximately 4.3 points/m2. The regression line between the results and the field data indicated the underestimation of tree height and crown diameter. Further work is needed to establish the influence of forest conditions and data with higher point densities.