目前常用的利用点云数据进行单木分割的方法大致可以分为两类：一种是基于归

At present, the commonly used methods of using point cloud data for single-wood segmentation can be roughly divided into two categories: one is based on normalized point cloud data for segmentation, called NPC method; the other is rasterization generated from point cloud data The canopy height model is based on single-wood segmentation. Common methods include watershed segmentation and polynomial fitting segmentation. The NPC method directly sets a reasonable threshold based on the spatial distance between the point clouds for clustering to complete the segmentation, while the method based on the canopy height model uses the height change of the forest canopy surface to detect tree vertices and crown edges. There are many image-based segmentation methods, such as edge detection algorithms, area segmentation, and image threshold segmentation.

目前常用的利用点云数据进行单木分割的方法大致可以分为两类：一种是基于归一化点云数据进行分割，称为NPC方法；另一种是由点云数据生成栅格化冠层高度模型，在此基础上进行单木分割，常用的方法有分水岭分割和多项式拟合分割等。NPC方法是直接基于点云之间的空间距离设定合理阈值进行聚类完成分割，而基于冠层高度模型的方法则是利用了森林冠层表面的高度变化进行树木顶点和树冠边缘的探测。基于图像的分割方法有很多，如边缘检测算法、区域分割、图像阈值分割等。

At present, the commonly used methods of single tree segmentation using point cloud data can be roughly divided into two categories: one is based on the normalized point cloud data, which is called NPC method; the other is based on the point cloud data to generate grid canopy height model, which is based on single tree segmentation, the commonly used methods are watershed segmentation and polynomial fitting segmentation. NPC method is based on the reasonable threshold set by the space distance between the point clouds to complete the clustering segmentation, while the method based on the canopy height model is based on the height change of the forest canopy surface to detect the tree vertex and crown edge. There are many image-based segmentation methods, such as edge detection algorithm, region segmentation, image threshold segmentation and so on.