A branch-trunk-constrained hierarchical clustering method for street trees individual extraction from mobile laser scanning point clouds

Abstract

The three-dimensional individual expression of street trees in the point cloud of urban road environments has important applications in many aspects. We propose a branch-trunk-constrained hierarchical clustering method to individually extract street trees from mobile laser scanning (MLS) point clouds. Firstly, potential tree crowns and trunks are selected through data preprocessing and regions of interest selection. Next, a branch-to-trunk matching method is proposed to locate the street tree positions. Then, tree crowns are segmented individually by a new hierarchical clustering method based on the identified tree positions. Finally, the tree trunk corresponding to each crown is selected from the potential ones to realize the complete individual extraction of street trees. The proposed method was tested and evaluated by three different MLS datasets. The results indicate that it can effectively extract individual street trees in different complex environments. Overall, the precision, recall and F-score of street tree positions locating have all reached 99.3%, and the precision, recall and F-score of street tree instance segmentation are 98.2%, 98.1% and 98.1%, respectively.