Abstract
Accurate individual tree segmentation is an important basis for the subsequent calculation and analysis of forestry parameters. However, rasterized canopy height model based methods often suffer from 3-D information loss due to the interpolation operation. Therefore, this article proposes an individual tree segmentation method based on the marker-controlled watershed algorithm and 3-D spatial distribution analysis from airborne LiDAR point clouds. First, based on the potential tree apices derived from the local maxima filtering, the marker-controlled watershed segmentation algorithm is conducted to obtain the coarse point clusters. Then, within the principal component analysis defined local coordinate reference framework, a multidirectional 3-D spatial profile analysis is performed on each point cluster to refine the potential tree apex positions. Finally, the refined potential tree apex positions are used as a prior of K-means clustering to achieve the coarse-to-fine individual tree segmentation. Comparative experiments were conducted on the public NEWFOR dataset to evaluate the proposed method. Results indicate that the proposed method is efficient and robust for segmenting individual trees.
Highlights
F OREST ecosystem, one of the largest and most important natural ecosystems among all the terrestrial ecosystems, is a natural complexity with specific structure, function, and self-regulation capability, formed by the forest community and its environment
Point cloud-based methods are increasingly becoming popular and can directly detect the individual tree crowns by analyzing their spatial distribution, which eliminates the effect of information loss due to rasterization compared with canopy height model (CHM)-based methods
2) Based on the coarse segmentation derived from CHM, the refined potential tree apex positions are used as a prior of K-means clustering algorithm to realize the coarse-to-fine segmentation of single trees
Summary
F OREST ecosystem, one of the largest and most important natural ecosystems among all the terrestrial ecosystems, is a natural complexity with specific structure, function, and self-regulation capability, formed by the forest community and its environment. Point cloud-based methods are increasingly becoming popular and can directly detect the individual tree crowns by analyzing their spatial distribution, which eliminates the effect of information loss due to rasterization compared with CHM-based methods. It is well known that point cloud-based method usually requires a lot of knowledge-based rules because of the unstructured nature of discrete point clouds, which might be not applicable To address these challenges, this article proposes a hybrid method, which combines the marker-controlled watershed algorithm and multidirectional 3-D spatial distribution analysis, to achieve a coarse-to-fine individual tree segmentation from airborne LiDAR point clouds.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have