期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2003
卷号：XXXIV-3/W13
出版社：Copernicus Publications
摘要：In recent years airborne laser scanning has been proven to be a helpful tool for remote sensing forestry applications. As laser scanning systems now provide very high point density (> 10 points/m 2 ), we pursue the approach of deriving geometric parameters on a single-tree basis. We explore the potential of delineating single trees from laser scanner raw data (x,y,z- triples) through cluster analysis and validate this approach with a dataset of about 2000 georeferenced trees. The dataset includes tree height and crown diameter and was gathered on a long term forest monitoring site by the Swiss Federal Institute for Forest, Snow and Landscape Research (WSL). A robust linear regression of field data tree heights with LIDAR derived tree heights leads to a slope of 0.91 and an offset of 0.71 m, with a RMS of 0.6 m. The accuracy of the laser scanner is evaluated through 6 reference targets, being 3x3 m 2 in size and horizontally plain. Internal offsets were found to be less than 0.25 m, the standard deviation of the points on the targets was about 0.06 m
关键词：Airborne laser scanning; segmentation; cluster analysis; point clouds; forestry
