【摘要】 提出了一种基于散乱点云的数据预处理方法.该方法包括四个部分:对散乱点云进行Dirichlet域分割并在此基础上进行三角剖分;在各个三角域中寻找中心点,以其为原点建立局部坐标系并采用正态分布模型进行噪声点删除;利用在三角网格上构建B-B曲面进行数据平滑处理;对漏测的数据点进行补全处理.数据点经过上述处理后能基本满足后续的曲面曲线的重构要求.更多还原

【Abstract】 A method of data preprocessing based on scattered point cloud was proposed.This method consisted of four parts: the scattered point cloud was segmented by Dirichlet method and then triangulated;the center point in each triangulation area was found and its local coordinate was set up,and a mathematical model of normal distribution was used to delete the noise points;further a B-B surface was made on each triangulation network to perform data smoothing processing;and finally,a supplementary processing was conducted for the missing data points.After processed in foregoing way,the data could basically meet the requirement of successive reconstruction of curved surfaces and curves.更多还原