Rapid Urban Roadside Tree Inventory Using a Mobile Laser Scanning System

Abstract

This paper presents a feasible workflow for use of three-dimensional point clouds acquired by a vehicle-borne mobile laser scanning (MLS) system for urban tree inventory. Extracting geometrical information, such as crown diameter, diameter at breast height (DBH), and tree height, from the MLS point clouds is a challenging task due to huge data volume, occlusions, mixed density, and irregular distribution of points in complex urban environments. The proposed workflow consists of three parts: individual tree cluster extraction, geometric parameter estimation, and tree species classification. The results show that over 93% of the roadside trees were correctly detected with an average error of about 5% in the DBH estimation when compared to field surveys and 78% of the overall accuracy was achieved for the classification of tree species.