基于点云处理的雪糕板锛头缺陷检测  

End Chips Defect Detection of Ice Cream Board Based on Point Cloud Processing

作　　者：刘月 白福忠[1] 高晓娟 李萍[3] 单文轩 LIU Yue;BAI Fu-zhong;GAO Xiao-juan;LI Ping;SHAN Wen-xuan(School of Mechanical Engineering,Inner Mongolia University of Technology,Hohhot010051,Inner Mongolia,P.R.China;School of Astronautics,Inner Mongolia University of Technology,Hohhot010051,Inner Mongolia,P.R.China;Beijing Polytechnic College,Beijing100042,P.R.China)

机构地区：[1]内蒙古工业大学机械工程学院,内蒙古呼和浩特010051 [2]内蒙古工业大学航空学院,内蒙古呼和浩特010051 [3]北京工业职业技术学院,北京100042

出　　处：《林产工业》2025年第3期75-81,共7页China Forest Products Industry

基　　金：内蒙古科技计划项目(2021GG0263);内蒙古自治区直属高校基本科研业务费项目(JY20220081);内蒙古自治区研究生科研创新项目(s20231123z)。

摘　　要：雪糕板锛头是一种三维缺陷,在生产中需要对其进行缺陷定位和定量检测以实现质量分级。本文提出了一种基于点云处理的雪糕板锛头缺陷检测方法。利用线结构光扫描获取雪糕板表面三维点云,通过迭代最近点(ICP)算法进行点云配准,实现了灰度图像辅助下的雪糕板点云区域分割。同时,采用随机采样一致性(RANSAC)与阈值判断相结合的算法进行锛头缺陷定位,进而实现缺陷的参数计算。结果表明:本方法与接触测量法结果相比,缺陷深度与长度的最大偏差分别仅为0.23 mm和0.12 mm,验证了该方法的有效性。The end chips of ice cream board are a three-dimensional(3D)defect.It is required for defect localization and quantitative detection in production to achieve quality classification.A defect detection method for end chips of ice cream board based on point cloud processing was proposed in this paper.3D point clouds of ice cream board surface were obtained by scanning with linear structured light.The iterative closest point(ICP)algorithm was used for point cloud registration,and the segmentation of ice cream board point cloud region was realized with the help of grayscale image.The algorithm of random sampling consistency(RANSAC)combined with threshold judgment was used to locate the end chips defect,and then the defect parameters were calculated.The experimental results showed that the maximum deviation of defect depth and length of the proposed method were only 0.23 mm and 0.12 mm compared with contact measuring technique,respectively.The effectiveness of the method was verified.

关 键 词：雪糕板 锛头缺陷 三维点云 点云配准 点云分割 

分 类 号：TS6[轻工技术与工程] TS396