PERSONAL LASER SCANNING TECHNOLOGY FOR 3D MAPPING AND FOREST INVENTORY

Abstract

LiDAR technology expands the possibilities for forest inventory using aerial and terrestrial scanning. A technological scheme for terrestrial personal laser scanning (PLS) has been developed in order to extract numerical tree parameters. The aim of the paper is to determine the characteristics of individual trees and average taxation indicators of forest stands via terrestrial PLS method. The PLS system includes Sensor Ouster OS1-16, mounted on a folding monopod and mini PC for control and recording of primary raw data, the base of creation of the 3D cloud of points. Google Cartographer is a system that provides simultaneous localization and 2D and 3D mapping. Cloud 3D model of the forest area was created by SLAM algorithm (Simultaneous Localization and Mapping). The result from SLAM is a 3D point cloud model in local coordinate system, which is transformed to WGS 84 (UTM 35N) coordinates via GPS data captured with LiDAR and IMU data. The 3D model is additionally processed by specialized application software Cloud Compare, LiDAR 360 and 3D Forest. An integrated algorithm is implemented through the software 3D Forest for segmentation of individual trees and determine their dendrometric parameters� height, diameter of the stem at breast height (DBH), structural data for the crown (diameter, length, height, area, volume), stem volume, and other. Study sites are forest stands on the territory of the South-Eastern Forest Enterprise (Sliven, Bulgaria). The sample plots are selected in accordance with the normative requirements for inventory and planning in forest territories of Bulgaria. The strip sampling method is based on a percentage system - limited proportion of the area is measured, on the assumption that the samples are typical of the entire stand. The obtained results for the individual trees are compared with the data from the parallel performed conventional field measurements at the sites. The differences between the measured by field methods and the PLS method tree height and stem diameter DBH of an individual trees and average data of a forest stands are within the normatively permissible limits in the forestry. Application of the PLS method combined with the SLAM algorithm provides the process of forest inventory and with the ability for fast, objective, and accurate measurement of individual trees and forest stands. The PLS method is a new reliable technological approach for 3D forest inventory and study of forest areas. Created 3D cloud models of forest stands are a qualitatively new type of objective data, which is a step towards to the future and creating of a new forestry database in Bulgaria.