[1]王立杰,高嘉伟,徐相龙,等. 基于SDA-YOLOv8的PCBA 缺陷检测算法研究[J].机械与电子,2026,44(01):45-51.
　WANG Lijie,GAO Jiawei,XU Xianglong,et al. Research on PCBA Defect Detection Algorithm Based on SDA -YOLOv8[J].Machinery & Electronics,2026,44(01):45-51.
点击复制

 基于SDA-YOLOv8的PCBA 缺陷检测算法研究()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
44
期数:
2026年01期
页码:
45-51
栏目:
智能检测
出版日期:
2026-01-27

文章信息/Info

Title:
 Research on PCBA Defect Detection Algorithm Based on SDA -YOLOv8
文章编号:
1001-2257(2026)01-0045-07
作者:
 王立杰; 高嘉伟; 徐相龙; 王婷婷
 (东北石油大学电气信息工程学院,黑龙江 大庆 163318)
Author(s):
WANG Lijie; GAO Jiawei; XU Xianglong; WANG Tingting
(School of Electrical and Infornation Engineering,Northeast Petroleum University,Daqing 163318,China)
关键词:
YOLOv8; 缺陷检测; PCBA; StarNet
Keywords:
YOLOv8; defect detection; PCBA; StarNet
分类号:
TP391.41
文献标志码:
A
摘要:
:针对电路板贴片元件缺陷检测中缺陷目标小、模型运算量大和训练生成模型文件过大问题,提出改进YOLOv8的轻量化缺陷检测算法。通过使用StarNet网络架构实现了高效的特征融合,降低计算量。设计了细节增强轻量化检测头,传递更多的低阶特征信息给高维检测网络,提高对小目标缺陷的检测效果。同时使用PRP AIFI模块替换原SPPF模块,更高效地捕捉有效信息点。实验数据表明,在自制数据集的测试实验中,改进后模型精度达到99.3%,该模型的mAP相比原模型提高了1.3%,模型参数量减少了40.9%,FLOPs最终为5.0×109,相较于原模型下降39%。所提算法在保持高精度的同时显著提升了运算效率。
Abstract:
To address the challenges of small defect targets,high computational load,and excessively large generative model files in the defect detection of surface mounted component on printed circuit boards,this paper proposes a lightweight defect detection method based on an improved YOLOv8.By adopting the StarNet architecture,efficient feature fusion is achieved with reduced computational cost.A detail enhanced lightweight detection head is designed to transmit more low level feature information to the high dimensional detection network,thereby improving the detection performance for small target defects.Additionally,the original SPPF module is replaced with the PRP AIFI module,enabling more effective capture of key information points.Experimental results on a custom dataset demonstrate that the improved model achieves a precision of 99.3%,with the mean Average Precision (mAP) increasing by 1.3% compared to the original model.The number of model parameters is reduced by 40.9%,and the FLOPs are reduced to 5.0×109,representing a 39%decrease relative to the baseline.The proposed algorithm significantly enhances computational efficiency while maintaining high accuracy.

参考文献/References:

[1] 杨溢青.基于深度学习的PCBA 焊点缺陷检测技术研究[D].桂林:桂林电子科技大学,2024.
[2] 沈明辉,刘宇杰,陈婧,等.基于改进YOLOv8s轻量化网络的组装电脑主板缺陷检测算法[J/OL].计算机工程,2025:1-14[2025-07-08].https:∥doi.org/10.19678/j.issn.1000 3428.0070196.
[3] 郑卓纹,吴攀超,王婷婷,等.基于轻量化目标检测算法的指针仪表读数识别[J].机械与电子,2025,43(5):10-17.
[4] 杨晨,章翔峰,姜宏.基于改进YOLOv8的施工安全装备检测算法[J].传感器与微系统,2025,44(6):138-142.
[5] KIM Y G,LIM D U,RYU J H,et al.SMD defect classification by convolution neural network and PCB image transform[C]∥2018 IEEE 3rd International Conference on Computing,Communication and Security(ICCCS).New York:IEEE,2018:180-183.
[6] CAI N,CEN G D,WU J X,et al.SMT solder joint inspection via a novel cascaded convolutional neural network[J].IEEE Transactions on components,packaging and manufacturing technology,2018,8(4):670-677.
[7] KIM Y G,PARK T H.SMT assembly inspection using dual stream convolutional networks and two solder regions[J].Applied sciences,2020,10(13):4598.
[8] 黄心玥,王明泉,耿宇杰,等.基于YOLOv8的铝合金铸造轮毂缺陷检测算法研究与分析[J].机械与电子,2025,43(2):26-31.
[9] MA X,DAI X Y,BAI Y,et al.Rewrite the stars[C]∥2014 IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York:IEEE,2024:5694-5703.
[10] ZHAO Y,LV W Y,XU S L,et al.Detrs beat yolos on real time object detection[C]∥2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition.New York:IEEE,2024:16965-16974.
[11] WU Y X,HE K M.Group normalization[C]∥Computer Vision – ECCV 2018.Cham:Springer,2018:3-19.
[12] CHEN Z X,HE Z W,LU Z M.DEA Net:single image dehazing based on detail enhanced convolution and content guided attention[J].IEEE Transactions on image processing,2024,33:1002-1015.
[13] 吴中凡,陆小锋,唐强达.基于改进YOLOv5s的复杂施工现场吸烟检测[J].计算机技术与发展,2024,34(10):31-37.
[14] 吴铁钰.基于改进YOLO 算法的道路缺陷识别研究[D].大连:大连交通大学,2025.
[15] 王鲜芳,刘依锋,杜志勇,等.深度神经网络蛋白质溶解性预测模型设计[J].河南师范大学学报(自然科学版),2021,49(2):31-39.

相似文献/References:

[1]陆 晓,高 超,蒋承伶,等.基于双注意力机制和级联检测框架的金具螺栓缺陷检测研究[J].机械与电子,2024,42(01):63.
　LU Xiao,GAO Chao,JIANG Chengling,et al.Research on Bolt Defect Detection Based on Dual-attention Mechanism and Cascade Detection Framework[J].Machinery & Electronics,2024,42(01):63.
[2]赖肖强,胡 泓.基于机器视觉算法和改进 MobileNetV2 的晶圆缺陷检测研究[J].机械与电子,2025,(01):34.
　LAI Xiaoqiang,HU Hong.Research on Wafer Defect Detection Based on Machine Vision Algorithms and Improved MobileNetV2[J].Machinery & Electronics,2025,(01):34.
[3]黄心玥,王明泉,耿宇杰,等.基于 YOLOv8 的铝合金铸造轮毂缺陷检测算法研究与分析[J].机械与电子,2025,(02):26.
　HUANG Xinyue,WANG Mingquan,GENG Yujie,et al.Research and Analysis of Defect Detection Algorithm for Aluminum Alloy Casting Wheels Based on YOLOv8[J].Machinery & Electronics,2025,(01):26.
[4]曹振锋,王明泉,路宇鹏,等.基于 YOLOv8 的轻量化轮毂缺陷检测算法研究[J].机械与电子,2025,(06):31.
　CAO Zhenfeng,WANG Mingquan,LU Yupeng,et al.Research on the Lightweight Wheel Hub Defect Detection Algorithm Based on YOLOv8[J].Machinery & Electronics,2025,(01):31.
[5]刘 骞,陈茂林.基于深度学习的少样本光伏边框划痕检测[J].机械与电子,2025,(08):47.
　LIU Qian,CHEN Maolin.Few-shot Scratch Detection on Photovoltaic Frame Based on Deep Learning[J].Machinery & Electronics,2025,(01):47.

备注/Memo

备注/Memo:
收稿日期:2025-08-06
基金项目:国家自然科学基金资助项目(52474036)
作者简介:王立杰 (1983-),男,山东黄县人,讲师,研究方向为先进控制算法、人工智能等;高嘉伟 (2000-),男,山东单县人,硕士研究生,研究方向为机器视觉,通信作者,E-mail:1942905616@qq.com;徐相龙 (2000-),男,黑龙江哈尔滨人,硕士研究生,研究方向为机器视觉;王婷婷 (1982-),女,黑龙江大庆人,博士,教授,博士研究生导师,研究方向为油气田岩石脆性分析。
更新日期/Last Update: 2026-03-09