结构光3D点云的PIN针针尖提取  

PIN tip extraction from 3D point cloud of structured light

作　　者：杜钦生[1] 李丹丹 陈浩 李雄飞[3,4] DU Qin-sheng;LI Dan-dan;CHEN Hao;LI Xiong-fei(College of Computer Science and Technology, Changchun University, Changchun 130022, China;DEEPerceptron Intelligent Technology Co., Ltd., Suzhou 215000, China;College of Computer Science and Technology, Jilin University, Changchun 130012, China;Engineering Institute, Tourism College of Changchun University, Changchun 130122, China)

机构地区：[1]长春大学计算机科学技术学院,吉林长春130022 [2]苏州深浅优视智能科技有限公司,江苏苏州215000 [3]吉林大学计算机科学与技术学院,吉林长春130012 [4]长春大学旅游学院工学院,吉林长春130122

出　　处：《液晶与显示》2021年第9期1331-1340,共10页Chinese Journal of Liquid Crystals and Displays

基　　金：吉林省教育厅科学研究规划项目(No.JJKH20200576KJ)。

摘　　要：在工业生产中对PIN针检测有越来越高的精度要求,针对点云数据存在毛刺现象、孔洞、离群点及大量不同类型的噪点等问题,本文提出采用结构光技术对PIN针的针尖平面在三维空间中进行提取的方法。首先,利用几何特性和直通滤波进行点云粗提取,快速准确地定位目标点云并去除大量非目标点云及离群点;然后,通过KD-tree对目标点云进行索引,使用欧氏距离聚类分割算法对点云数量进行分割,稳定有效地去除目标点云附近的小范围噪点;最后,通过对目标点云法向量与基准面法向量夹角的判断方法有效且精确地去除目标点云内不平整的噪点。实验结果表明,该方法不仅可以准确去除PIN针点云数据的噪点,且能精确地提取针尖坐标,不同方向的PIN针高度测量标准偏差在0.005 mm以内。本文提出的方法普遍适用于平头型PIN针针尖的提取,精度高,速度快,鲁棒性能好。In industrial production,there is a higher precision requirement for pin detection.In view of the burr phenomenon,holes,outliers and a large number of different types of noise in point cloud data,this paper proposes a method to extract the pin tip plane in three-dimensional space using structured light technology.First of all,rough extraction of point cloud is carried out by geometric characteristics and through filtering.It can locate the target point cloud quickly and remove a large number of non-target point clouds and outliers accurately.Then,the target point cloud is indexed by KD-tree,the Euclidean distance clustering segmentation algorithm is used to segment the number of point clouds.It can stably and effectively remove small-scale noise near the target point cloud.Finally,the angle is judged between the vector of the target point cloud and the normal vector of the reference plane.It can remove the uneven noise in the target point cloud effectively and accurately.The experimental results show that this method can not only remove the noise of the pin point cloud data accurately,but also extract the pin point coordinates accurately.The standard deviation of the pin height measurement in different directions is within 0.005 mm.The method proposed in this paper is generally applicable to the extraction of flat pin tip with high precision,high speed and good robustness.

关 键 词：PIN针 结构光 点云分割 点云滤波 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]