基于水平截面法的树木点云精细分割研究

Research on Fine Segmentation of Tree Point Clouds Based on Horizontal Section Method

本文提出了水平截面法精细分割树木点云的方法。首先根据点云数据在XOY平面内建立二维格网,求取每个格网内点云数据高度的均值,根据格网内点云数据的高度阈值滤除地面点云和低矮地物点云数据;其次对滤除地面和低矮地物的点云运用八邻域算法提取树木点云,运用水平截面法对树木分层处理,求取每一层树木点云的轮廓线;再次将轮廓线连成多边形,求取多边形的最小外接圆的圆心,运用最小二乘直线拟合算法对各层圆心进行直线拟合,根据拟合直线与地面的交点,可以求取树木的精确位置;最后根据距离最近的原理,对树木进行了精细分割。实验表明:本文研究的方法可以较好的实现树木点云的提取,避免了传统方法设置参数过多的弊端。

This paper proposes a method of finely segmenting tree point clouds by horizontal section method.Firstly,according to the point clouds,a two-dimensional grid is established in the XOY plane,and the mean value of the point clouds height in each grid is obtained.according to the height threshold of point clouds data in the grid,ground point clouds and lowland point clouds are Filtered;Secondly,ground and low ground objects are filtered,and thentree point clouds are extracted by eight-neighbor algorithm,and the horizontal section method are used to layer the trees and then obtain the outline of the point clouds of each layer of trees;thirdly,Concatenate the outlines into polygons again and find the center of the smallest circumscribed circle of the polygon.The least squares straight line fitting algorithm is used to fit the center of each layer.According to the intersection of the fitted line and the ground,position of the tree can be obtained.Finally,according to the nearest principle,the trees are finely divided.Experiments show that the method studied in this paper can better realize the extraction of tree point clouds and avoid the disadvantages of too many parameters in traditional methods.