基于扇区邻域特征工程的玻璃封装绝缘端子缺陷检测  

Defect Detection for Glass Seal Insulated Terminals Based on Sector Neighborhood Feature Engineering

作　　者：蔡念[1,4] 李炜博 黄钦豪 周帅 邱宝军[3] 何兆泉[1] CAI Nian;LI Weibo;HUANG Qinhao;ZHOU Shuai;QIU Baojun;HE Zhaoquan(School of Information Engineering,Guangdong University of Technology,Guangzhou 510006,China;Tianjin University,Tianjin 300072,China;China Electronic Product Reliability and Environmental Testing Research Institute,Guangzhou 510006,China;Huizhou Guangdong University of Technology IoT Cooperative Innovation Institute Co.,Ltd.,Huizhou 516025,China)

机构地区：[1]广东工业大学信息工程学院,广州510006 [2]天津大学,天津300072 [3]工业和信息化部电子第五研究所,广州510006 [4]惠州市广工大物联网协同创新研究院有限公司,惠州516025

出　　处：《电子与信息学报》2022年第5期1548-1553,共6页Journal of Electronics & Information Technology

基　　金：国家自然科学基金(62171142);广东省自然科学基金(2021A1515011908);惠州市高校科研专项资金(2019HZKY003)。

摘　　要：该文提出一种基于特征工程的玻璃封装绝缘端子外观质量检测算法以替代人工肉眼检测。首先,基于绝缘端子的形状先验将待检测区域划分为若干个扇区。其次,根据玻璃封装绝缘端子图像特点提出扇区基本特征数据、扇区灰度变化率、扇区反光特征和扇区方向统计特征等4大类扇区特征提取方法,采用梯度提升决策树(GBDT)进行粗分类。为了更全面地表征扇区特性,基于粗分类结果融合最近邻扇区提出扇区近邻(SN)特征提取方法。最后,将扇区近邻特征和扇区特征都输入到GBDT分类器进行精细分类,获得检测结果。实验结果表明,提出的特征工程方法能够在合理检测时间内取得较好的检测性能,交并比为97.45%,F1为0.987,优于现有类似检测方法。In this paper,an appearance quality inspection method for glass seal insulated terminals is proposed based on feature engineering to replace the current manual inspection.First,the inspecting region in the glass seal insulated terminal image is divided into many sectors based on the shape prior of the terminal.Second,considering the characteristics of the image,four categories of sector features are designed,such as sector basic features,sector gray change rates,sector reflection features and sector direction statistical features.Then,they are input into a Gradient Boosting Decision Tree(GBDT)for rough classification.Next,to characterize excellently the sectors,a new image feature called Sector Neighborhood(SN)feature is designed by combining sector features of the nearest neighbor sectors and rough classification results for these sectors.Finally,the SN features and sector features are input into the GBDT for fine classification,which indicates final quality inspection.The experimental results indicate that the proposed method can achieve better inspection performance with reasonable inspection time compared to the existing inspection methods,which has 97.45%IoU and 0.987 F1.

关 键 词：外观质量检测 玻璃封装绝缘端子 扇区特征 扇区近邻特征 梯度提升决策树 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] TP391.4[自动化与计算机技术—控制科学与工程]