[1]明月,吕清花,翟中生,等.基于改进YOLOv5算法的晶圆表面缺陷检测方法[J].湖北工业大学学报,2024,39(4):98-105.
　MING Yue,LV Qinghua,ZHAI Zhongsheng,et al.Wafer Surface Defect Detection Method Basedon Improved YOLOv5 Algorithm[J].,2024,39(4):98-105.
点击复制

基于改进YOLOv5算法的晶圆表面缺陷检测方法()
分享到：

《湖北工业大学学报》[ISSN:1003-4684/CN:42-1752/Z]

卷:
39
期数:
2024年第4期
页码:
98-105
栏目:
出版日期:
2024-08-28

文章信息/Info

Title:
Wafer Surface Defect Detection Method Basedon Improved YOLOv5 Algorithm
文章编号:
1003-4684(2024)04-0098-08
作者:
明月1;  吕清花1;  翟中生2;  吕辉1;  於意凯1;  崔贤岱1
(1 湖北工业大学理学院， 湖北 武汉 430070； 2 湖北工业大学机械工程学院， 湖北 武汉 430070)
Author(s):
MING Yue1;  LV Qinghua1;  ZHAI Zhongsheng2;  LV Hui1;  YU Yikai1;  CUI Xiandai1
(1 School of science, Hubei Univ. of Tech., Wuhan 430070, China;2 School of Mechanical Engineering, Hubei Univ. of Tech., Wuhan 430070, China)
关键词:
深度学习;  晶圆表面缺陷;  缺陷检测;  YOLOv5;  GhostNet
Keywords:
deep learning;  wafer surface defects;  defect detection;  YOLOv5;  GhostNet
分类号:
TP391.4, TN407
文献标志码:
A
摘要:
为了兼顾实时性和准确率，提出了一种基于改进YOLOv5算法的晶圆表面缺陷检测方法。该方法采用了轻量级网络GhostNet作为主干提取网络，以降低模型复杂度并提升检测速度。同时为了提高模型的特征提取能力和检测精度，引入了高效通道注意力机制。此外采用FReLU激活函数取代了原有的SiLU函数，以增强模型对空间的敏感性，提高检测准确性。使用真实的晶圆缺陷数据集对改进模型进行验证。实验结果表明，相比于原始模型，改进YOLOv5网络模型实现了30.02%的参数压缩，同时目标精度达到78.6%，相较于YOLOv5s提升了4.4%，mAP值提高5.5%，检测速度提高1.3 ms。
Abstract:
Wafer surface defect detection holds significant importance in semiconductor chip manufacturing. However, during the inspection process, false detection and missed detection of defects often occur due to the complexity and diversity of wafer surface defect types and manifestations. To balance real time and accuracy requirements, a wafer surface defect detection method based on the improved YOLOv5 algorithm is proposed. This method uses the lightweight network GhostNet as the backbone extraction network to reduce model complexity and improve detection speed. Additionally, an efficient channel attention mechanism is introduced to enhance the model’s feature extraction ability and detection accuracy. The original SiLU function is replaced with the FReLU activation function to improve the model’s sensitivity to space and detection accuracy. The improved model is validated using a real wafer defect dataset. The experimental results show that the improved YOLOv5 network model achieves 30.02% parameter compression compared with the original model. The target accuracy reaches 78.6%, which is 4.4% higher than YOLOv5s. The mAP value is increased by 5.5%, and the detection speed is increased by 1.3 ms.

参考文献/References:

［1］KONG Q, WU Z, SONG Y. Online detection of external thread surface defects based on an improved template matching algorithm [J ]. Measurement, 2022, 195: 111087. [2 ]JOHN V, YONEDA K, LIU Z, et al. Saliency map generation by the convolutional neural network for real -time traffic light detection using template matching [J ]. IEEE transactions on computational imaging, 2015, 1(03): 159 -173. [3 ]YU X K,WANG Z W,WANG Y H,et al. Edge detection of agricultural products based on morphologically improved canny algorithm [J ]. Mathematical Problems in Engineering, 2021, 2021(03):1 -10. [4 ]王睿男,武穆清,陈铁英,等.基于形态学的红外图像边缘检测 [J ].北京邮电大学学报,2021,44(01):66 -71. [5 ]曹宇, 徐传鹏. 一种改进阈值分割算法在镜片缺陷检测中的应用 [J ].激光与光电子学进展,2021, 58(16):219 -224. [6 ]周晓彦,王珂,李凌燕. 基于深度学习的目标检测算法综述 [J ]. 电子测量技术,2017,40(11):89 -93. [7 ]杨传礼,张修庆.基于机器视觉和深度学习的材料缺陷检测应用综述 [J ].材料导报,2022,36(16):226 -234. [8 ]靳晶晶,王佩.基于卷积神经网络的图像识别算法研究 [J ].通信与信息技术,2022(02):76 -81. [9 ]张上,王恒涛,冉秀康. 基于YOLOv5的轻量化交通标志检测方法 [J ]. 电子测量技术,2022,45(08):129 -135. [10 ] 张立国,马子荐,金梅,等.基于YOLO的轻量红外图像行人检测方法 [J ].激光与红外,2022,52(11):1737 -1744. [11 ] XIANG Y, FEI Q. Car, Cyclist and pedestrian object detection based on YOLOv5 [J ]. International Journal of New Developments in Engineering and Society, 2022, 6(03)：38 -42. [12 ] FANG C, XIANG H, LENG C, et al. Research on real -time detection of safety harness wearing of workshop personnel based on YOLOv5 and OpenPose [J ]. Sustainability, 2022, 14(10): 5872. [13 ] CAO X, ZHANG F, YI C, et al. Wafer surface defect detection based on improved YOLOv3 network [C ]. Proceedings -2020 5th International Conference on Mechanical, Control and Computer Engineering, ICMCCE 2020, 9421681 :1475 -1478. [14 ] XIE Y, HU W, XIE S, et al. Surface defect detection algorithm based on feature -enhanced YOLO [J ]. Cognitive Computation, 2022: 1 -15. [15 ] CHEN S H, TSAI C C. SMD LED chips defect detection using a YOLOv3 -dense model [J ]. Advanced engineering informatics, 2021, 47: 101255. [16 ] ZHANG D,CHEN X,REN Y,et al．SmartGYOLO: alight -weight real -time object detection network [J ]．Journal of Physics: Conference Series,2021, 1757: 12096. [17 ] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: Unified, real -time object detection [C ].Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 779 -788. [18 ] HAN K, WANG Y, TIAN Q, et al. Ghostnet: More features from cheap operations [C ] Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. 2020: 1577 -1586. [19 ] HU J, SHEN L, ALBANIE S. et al. Squeeze and excitation networks [J ]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2020,42(08):2011 -2023. [20 ] LIU J W, LIU J W, LUO X L. Research progress in attention mechanism in deep learning [J ]. Gongcheng Kexue Xuebao/Chinese Journal of Engineering, 2021, 43(11): 1499 -1511. [21 ] WANG Q, WU B, ZHU P, et al. ECA -Net: Efficient channel attention for deep convolutional neural networks [C ]. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. 2020: 11531 -11539. [22 ] WOO S, PARK J, LEE J Y, et al. Cbam: Convolutional block attention module [C ]. Proceedings of the European conference on computer vision (ECCV). 2018: 3 -19. [23 ] FU J, LIU J, TIAN H, et al. Dual attention network for scene segmentation [C ]. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2019: 3146 -3154. [24 ] SHIV R D, SATISH K S, BIDYUT B C. Activation functions in deep learning: A comprehensive survey and benchmark [J ]. Neurocomputing, 2022,503(07):92 -108. [25 ] MA N, ZHANG X, SUN J. Funnel activation for visual recognition [C ]. Computer Vision -ECCV 2020: 16th European Conference, Glasgow, UK, August 23 -28, 2020, Proceedings, Part XI 16. Springer International Publishing, 2020: 351 -368.

相似文献/References:

[1]周 浩,周先军,邱书畅.下采样迭代和超分辨率重建的图像风格迁移[J].湖北工业大学学报,2020,(1):25.
　ZHOU Hao,ZHOU Xianjun,QIU Shuchang.Image Style Transfer Algorithm Based on Downsampling Iterative and Super-Resolution Reconstruction[J].,2020,(4):25.
[2]张 宇,程 玉,陈建峡,等. 基于深度学习的在线推荐学习系统设计与开发[J].湖北工业大学学报,2021,(5):64.
　ZHANG Yu,CHENG Yu,CHEN Jianxia,et al. Design and Development of Online Recommendation Learning System Based on Deep Learning[J].,2021,(4):64.
[3]武明虎,雷常鼎,刘 聪. 改进的轻量级YOLOv4输电线路鸟巢检测方法[J].湖北工业大学学报,2023,(2):11.
　WU Minghu,LEI Changding,LIU Cong. Improved YOLOv4 Transmission Lines Bird’s Nest Detection Method[J].,2023,(4):11.
[4]董　庚１,王焱清１,２,等.基于 ResNet１８网络的油茶果壳籽分选研究[J].湖北工业大学学报,2023,(5):29.
　DONG Gen,WANG Yanqing,SUN Jiwei,et al.Sorting Model of Camellia Fruit Shells and TeaSeeds Based on ResNet18 Network[J].,2023,(4):29.

备注/Memo

备注/Memo:
［收稿日期］ 20230407［基金项目］ 武汉市重点研发计划（2022012202015034）［第一作者］ 明月（1996-）， 女， 湖北孝感人， 湖北工业大学硕士研究生， 研究方向为集成电路检测。［通信作者］ 吕清花（1980-）， 女， 湖北咸宁人， 湖北工业大学副教授， 研究方向为集成电路、 新型光电器件、 光电检测等。
更新日期/Last Update: 2024-08-23