Individual Tree Reconstruction Based on Circular Truncated Cones From Portable LiDAR Scanner Data

Abstract

The LiDAR scanning approach captures a large amount of tree point cloud information, including the accurate coordinate of points, local topology, and overall geometry. However, the complex tree structures, e.g., curvature and occlusion of branches, bring challenges to the 3-D tree reconstruction. In this letter, we propose an innovative solution for obtaining the complete skeletons of individual trees and 3-D structures for modeling using point clouds. First, we obtain individual trees from input street scene and segment tree into small successive pieces, with the centers of each piece serving as skeleton candidate points. Second, we conduct the interpolation based on the Euclidean distance and orientation yields the entire skeleton completely. Finally, a high-precision 3-D model of trees is constructed by cylindrically fitting the skeleton relying on the optimized circular truncated cones depending on the branch orientation and cylindrical curvature. Experiments on various trees demonstrate the high efficiency and effectiveness of our method. Our method achieves 98% accuracy and takes less than 1 min in the reconstruction. Compared with other methods, our method reduces the time by more than 95%.