An Automatic Tree Skeleton Extracting Method Based on Point Cloud of Terrestrial Laser Scanner

Abstract

Tree skeleton could describe the shape and topological structure of a tree, which are useful to forest researchers. Terrestrial laser scanner (TLS) can scan trees with high accuracy and speed to acquire the point cloud data, which could be used to extract tree skeletons. An adaptive extracting method of tree skeleton based on the point cloud data of TLS was proposed in this paper. The point cloud data were segmented by artificial filtration and k-means clustering, and the point cloud data of trunk and branches remained to extract skeleton. Then the skeleton nodes were calculated by using breadth first search (BFS) method, quantifying method, and clustering method. Based on their connectivity, the skeleton nodes were connected to generate the tree skeleton, which would be smoothed by using Laplace smoothing method. In this paper, the point cloud data of a toona tree and peach tree were used to test the proposed method and for comparing the proposed method with the shortest path method to illustrate the robustness and superiority of the method. The experimental results showed that the shape of tree skeleton extracted was consistent with the real tree, which showed the method proposed in the paper is effective and feasible.