【摘要】 散乱点云离群点识别和表面平滑作为点云预处理的主要组成部分,是三维场景建模和可视化的重要前提。针对这一问题,论文提出了基于随机测量误差特点和噪声点分布特性的点云降噪方法:离群点识别采用统计分类的思想,通过特征提取将点云映射到特征空间后加以区分;表面平滑利用噪声点分布特性,对采样点的真实位置进行估计。实验结果表明,采用文中的方法能够准确有效地识别离群点和平滑模型表面。更多还原

【Abstract】 As main part of the preprocessing of scattered point cloud,outlier identification and surface smoothing are important premise of 3Dmodeling and visualization.In order to solve this problem,some certain effective denoising methods are put forward based on the characteristics of random measurement error and noise distribution of point cloud:identifying outliers by mapping the point cloud into feature space through feature extraction;surface smoothing estimates the true locations of sampling points based on the noise distribution.The experimental results indicate that the proposed methods can identify outliers accurately and smooth model surface effectively.更多还原