摘要：

针对点云模型分割和特征面识别速度慢、准确性差的问题,提出基于连通区域标记和统计法的散乱点云特征面分割与识别方法。通过估算点云法矢与点云曲率,给出零值固定曲率归一化方法。基于曲率对点云初始聚类,采用连通区域标记法分割点云,进而利用统计法判断点云所属特征类型曲面。实验结果表明,在以规则曲面为主的机械零件特征曲面分割和识别应用中,该方法能够满足中小型规模的点云处理需求。

关键词: 点云分割, 法矢估计, 曲率估算, 曲面识别, 特征曲面

Abstract: Aiming at the problem of slow speed and poor accuracy of model segmentation and feature surface recognition,a point cloud segmentation and surface recognition method based on connected component labeling and probabilistic method is proposed.A zero-value fixed curvature normalization method is proposed by estimating the curvature of the point cloud and the curvature of the point cloud.Based on the initial clustering of curvature on the point cloud,the connected region marker method is used to segment the point cloud,and then the statistical method is used to judge the feature type surface of the point cloud.Experimental results show that the proposed method can meet the demand for processing small and medium size point clouds in the segmentation and application of feature surface of mechanical parts based on regular surfaces.

Key words: point cloud segmentation, normal estimation, curvature calculation, surface recognition, feature surface

中图分类号: 

• TP399