基于改进YOLOv5的陶瓷基片缺陷识别  

Defect identification on ceramic substrates based on improved YOLOv5

作　　者：王鹏飞 李海洋[2] 廖健标 林骁 Wang Pengfei;Li Haiyang;Liao Jianbiao;Lin Xiao(College of Information and Communication Engineer,North University of China,Taiyuan 030051,China;Shanghai Acoustics Laboratory,Chinese Academy of Sciences,Shanghai 201815,China;Zhejiang Xinna Material Technology Co.,Ltd.,Jinhua 322118,China)

机构地区：[1]中北大学信息与通信工程学院,太原030051 [2]中国科学院声学研究所东海研究站,上海201815 [3]浙江新纳材料科技股份有限公司,金华322118

出　　处：《国外电子测量技术》2024年第11期187-196,共10页Foreign Electronic Measurement Technology

基　　金：山西省基础研究计划(202203021221099);国家重点研发计划(2023YFF0720904)项目资助。

摘　　要：陶瓷基片的缺陷严重影响电子器件的性能,为提高缺陷检测的准确性,基于超声显微镜扫描的陶瓷基片检测方法,提出了一种改进YOLOv5的神经网络算法。根据超声检测具有穿透性的优点,增加一条新的主干网络综合陶瓷基片表面与内部的回波信息,同时使用极化注意力机制进行特征融合提高检测的精确度,并融合了轻量化网络减少参数量。进行了超声显微镜扫描陶瓷基片实验分析缺陷特征并制作数据集,在此数据集上,FusionPol-YOLOv5模型对9种缺陷检测的精确率达到88.3%,平均精度均值(mAP)mAP@0.5达到91.7%,可以极大减少陶瓷基片检测的人力物力损耗和成本。The defects of ceramic substrates have a significant impact on the performance of electronic devices.To enhance the accuracy of defect detection,in this paper,based on the detection method of ceramic substrates by ultrasonic microscopy scanning,an improved neural network algorithm of YOLOv5 is proposed.Taking advantage of the penetrability of ultrasonic detection,a new backbone network is added to comprehensively integrate the echo information from both the surface and interior of the ceramic substrates.Meanwhile,a polarization attention mechanism is employed for feature fusion to improve the detection precision,and a lightweight network is integrated to reduce the number of parameters.Experiments of ultrasonic microscopy scanning on ceramic substrates were carried out to analyze the defect characteristics and create a dataset.On this dataset,the FusionPol-YOLOv5 model proposed in this paper achieves an precision of 88.3%for the detection of 9 types of defects,with an mAP@0.5 of 91.7%.It can significantly reduce the human and material resources consumption and costs in the detection of ceramic substrates.

关 键 词：缺陷识别 超声检测 注意力机制 特征融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TN04[自动化与计算机技术—计算机科学与技术]