摘要
为了实现导弹发射箱的自动化打磨工序,必须提取复杂箱体的特征尺寸。通过三维激光扫描技术得到发射箱的点云数据后,对其进行分析,提出了一种基于法向量和排序欧式聚类(Normal Vector and Sorted Euclidean Clustering,NV-SEC)的箱体类零件特征尺寸快速提取方法。该方法首先以箱体点云模型中各点法向量的不同为判断依据,将箱体所有的环筋边缘信息提取出来,并采用改进的排序欧式聚类分割算法,得到有序的单根环筋的边缘信息,从而进一步求得箱体的宽度尺寸。实验结果表明,该方法可以满足箱体自动化打磨的要求,可以大幅度提高箱体环筋与纵筋装配质量,提高箱体加工效率。
In order to achieve automatic grinding of the missile launching box, the feature size of the complex box must be extracted. After obtaining the point cloud data of the launching box by 3 D laser scanning technology, a fast extraction algorithm based on normal vector and sorted Euclidean clustering(NV-SEC) is proposed in this study to obtain the feature size of the box parts. Based on the difference of the normal vectors of each point in the box point cloud model, the algorithm extracts the whole edge information of the ring at first. Then the improved Euclidean clustering is used to fetch the edge information of the sorted single ring and further the width dimension of the box is obtained. The experimental results show that the algorithm can meet the requirements of automatic grinding of the box, which can greatly improve the assembly quality of the box ring and longitudinal reinforcement and improve the processing efficiency of the box.
作者
秦立功
李蓓智
黄顺舟
杨建国
QIN Li-gong;LI Bei-zhi;HUANG Shun-zhou;YANG Jian-guo(College of Mechanical Engineering,Donghua University,Shanghai 201600,China;Shanghai Aerospace Equipment Manufacturing General Co.,Ltd.,Shanghai 201100,China)
出处
《组合机床与自动化加工技术》
北大核心
2020年第1期51-53,58,共4页
Modular Machine Tool & Automatic Manufacturing Technique
基金
高档数控机床与基础制造装备(2017ZX04005001)
关键词
三维点云
法向量
尺寸提取
聚类分割
3D point cloud
normal vector
size extraction
cluster segmentation
