基于支持向量机的印板缺陷检测  

Defect detection of printed board based on support vector machine

作　　者：胡洋 韩加兴 石煌雄 侯俊龙 蒋作[3] 潘文林[3] HU Yang;HAN Jia-xing;SHI Huang-xiong;HOU Jun-long;JIANG ZuoPAN Wen-lin(School of Electrical Information Engineering,YunnanMinzuUniversity,Kunming 650500,China;Doctor Octopus Intelligent Technology(Shanghai)Co.,Ltd,Shanghai 201800,China;School of Mathematics and Computer Science,YunnanMinzuUniversity,Kunming 650500,China)

机构地区：[1]云南民族大学电气信息工程学院,云南昆明650500 [2]章鱼博士智能技术(上海)有限公司,上海嘉定区201800 [3]云南民族大学数学与计算机科学学院,云南昆明650500

出　　处：《云南民族大学学报（自然科学版）》2023年第1期124-130,共7页Journal of Yunnan Minzu University:Natural Sciences Edition

基　　金：云南省教育厅科学研究基金(2020Y0237).

摘　　要：印板在工业制造中会产生表面缺陷,对产品的外观和性能都有影响.目前没有针对印板缺陷检测相关产品,为了解决这一问题,设计一个系统能够自动检测印板缺陷.根据采集的样本,确定缺陷种类.提取印板缺陷的纹理、几何和Hu不变矩特征共计23种,确定15种特征作为印板缺陷分类的依据,并运用支持向量机(SVM)方法建模.经实验验证,该系统能对印板常见3类缺陷检测,准确率高,检测速度平均0.7 s/张.Printed board will produce surface defects in industrial manufacturing,which will affect the appearance and performance of the product.At present,there is no product related to the detection of printed board defects.To solve this problem,a system is designed to automatically detect printed board defects.According to the collected sample data,the type of defect was determined.A total of 23 kinds of texture,geometry and Hu invariant moment features of printed board defects were extracted,15 kinds of features were determined as the basis for classification of printed board defects,and a support vector machine(SVM)method was used for modeling.It has been verified by experiments that the system can detect three common types of defects on printed boards,with high accuracy and an average detection speed of 50 ms/sheet.

关 键 词：缺陷检测 纹理特征 几何特征 HU不变矩 支持向量机 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]