摘要
点云数据分区是逆向工程中重要而又难以解决的问题。首次将模糊聚类方法应用于逆向工程中的点云数据分区,用点的位置矢量、法矢量、高斯曲率和平均曲率8维向量作为特征向量,加权距离替代欧氏距离。在实现分区的同时,可以识别区域内部点和边界附近点,便于后续曲面特征参数精确提取。实验结果证明此算法具有较强的抗噪性,并具有较高的分区效率。
Point cloud data segmentation is an important but difficult question in reverse engineering. For the first time, the fuzzy c-means clustering algorithm was applied to the point cloud data segmentation. 8D feature vectors of points including 3D coordinates, 3D normal vector, mean curvature and Gauss curvature were taken as input feature vectors, and weighted distance replaced the Euclidean distance. The algorithm can also identify inner points and border points at the same time when the segmentation was implemented, creating convenience for extracting accu- rately the feature parameters of subsequent surfaces. Experimental results show that the algorithm has strong noise resistance and efficient segmentation.
出处
《机械科学与技术》
CSCD
北大核心
2007年第4期515-517,520,共4页
Mechanical Science and Technology for Aerospace Engineering
基金
国家自然科学基金项目(60573177)
航空科学基金项目(04H53059)
河南省教育厅自然科学基金项目(200510078010)资助
关键词
模糊聚类
逆向工程
点云分区
fuzzy c-means clustering algorithm
reverse engineering
point cloud data segmentation
