Adaptive clustering of airborne LiDAR data to segment individual tree crowns in managed pine forests

• Published in: International journal of remote sensing Vol. 31; no. 1; pp. 117 - 139
• Main Authors: Lee, H., Slatton, K. C., Roth, B. E., Cropper, W. P.
• Format: Journal Article
• Language: English
• Published: Abingdon Taylor & Francis 01.03.2010
• Subjects: Animal, plant and microbial ecology; Applied geophysics; Biological and medical sciences; Canopies; Clustering; Earth sciences; Earth, ocean, space; Exact sciences and technology; Forests; Fundamental and applied biological sciences. Psychology; General aspects. Techniques; Internal geophysics; Lidar; Pine; Search methods; Teledetection and vegetation maps; Trees; Central Florida; Florida; United States; data; Irregularity; Size; adaptation; measurement sensor; Determination; clouds; Pinus; Image; Information; North America; trees; Result; Pinaceae; variations; Tree crown; Three dimensional system; segmentation; gymnosperms; seeds; Plantae; utilization; Method; windows; Individual; Lidar; Parameter; Coniferales; Spermatophyta; Canopy(vegetation); Coniferous forest
• ISSN: 0143-1161; 1366-5901
• DOI: 10.1080/01431160902882561

Measuring individual trees can provide valuable information about forests, and airborne light detection and ranging (LiDAR) sensors have been used recently to identify individual trees and measure structural tree parameters. Past results, however, have been mixed because of reliance on interpolated (image) versions of the LiDAR measurements and search methods that do not adapt to variations in canopies. In this work, an adaptive clustering method is developed using airborne LiDAR data acquired over two distinctly different managed pine forests in North-Central Florida, USA. A crucial issue in isolating individual trees is determining the appropriate size of the moving window (search radius) when locating seed points. The proposed approach works directly on the three-dimensional (3D) 'cloud' of LiDAR points and adapts to irregular canopies sizes. The region growing step yields collectively exhaustive sets in an initial segmentation of tree canopies. An agglomerative clustering step is then used to merge clusters that represent parts of whole canopies using locally varying height distribution. The overall tree detection accuracy achieved is 95.1% with no significant bias. The tree detection enables subsequent estimation of tree height and vertical crown length to an accuracy better than 0.8 and 1.5 m, respectively.