基于数据挖掘技术的牙刷包装缺陷检测方法研究  

Research on Defect Detection Method of Toothbrush Packaging Based on Data Mining Technology

作　　者：刘贺[1] 马小燕[2] 张静[3] LIU He;MA Xiao-yan;ZHANG Jing(Information Center of Yangzhou Polytechnic Institute, Yangzhou Jiangsu 225127,China;School of Intelligent Manufacturing, Yangzhou Polytechnic Institute, Yangzhou Jiangsu 225127,China;School of Information Engineering, Yangzhou Polytechnic Institute, Yangzhou Jiangsu 225127,China)

机构地区：[1]扬州工业职业技术学院信息中心,江苏扬州225127 [2]扬州工业职业技术学院智能制造学院,江苏扬州225127 [3]扬州工业职业技术学院信息工程学院,江苏扬州225127

出　　处：《佳木斯大学学报（自然科学版）》2020年第3期48-51,共4页Journal of Jiamusi University：Natural Science Edition

基　　金：2018年江苏省市校合作项目(YZ2018147)。

摘　　要：为了实现对牙刷包装缺陷的精准检测,提出基于数据挖掘技术的牙刷包装缺陷检测方法。构建牙刷包装视觉信息采样模型,对牙刷包装视觉图像进行超像素特征提取。建立牙刷包装缺陷的大数据分布集,结合模糊信息采样技术进行牙刷包装缺陷的信息采样和多分辨融合,建立牙刷包装图像的信息融合模型。采用统计分析方法进行牙刷包装缺陷分布特征检测,根据信息增强和模糊聚类分析方法,实现牙刷包装缺陷检测。仿真结果表明,采用该方法进行牙刷包装缺陷检测的精度较高,自适应性较好,提高了牙刷包装缺陷定位和特征提取能力。In order to realize accurate detection of toothbrush packaging defects,a method based on data mining technology was proposed.The visual information sampling model of toothbrush packaging was established to extract the super pixel features of the visual image of toothbrush packaging.The big data distribution set of toothbrush packaging defects was established,the information sampling and multi-resolution fusion of toothbrush packaging defects were carried out by combining the fuzzy information sampling technology,and the information fusion model of toothbrush packaging image was established.Statistical analysis method was used to detect the distribution characteristics of toothbrush packaging defects,and based on information enhancement and fuzzy cluster analysis method,the defect detection of toothbrush packaging was realized.The simulation results show that this method has high accuracy and good adaptability in detecting the defects of toothbrush packaging,and improves the ability of locating the defects and extracting the features.

关 键 词：数据挖掘 牙刷 包装 缺陷 检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]