基于改进YOLOv5s的PCB缺陷检测算法

doi:10.19678/j.issn.1000-3428.0070376

摘要/Abstract

摘要：

印刷电路板(PCB)的质量对工业电子产品的性能具有决定性的作用, 因此对PCB出厂质量的严格把控具有重要意义。PCB易出现的6类缺陷是评价PCB质量的主要参考依据, 针对现有的目标检测算法模型对PCB缺陷的检测精度低、体积庞大等问题, 提出改进的YOLOv5s-CMS模型。根据PCB缺陷的特点, YOLOv5s-CMS模型采用聚焦小目标信息的特征提取网络CCG(C2f-C3-Ghost)替换原有的特征提取网络, 使模型在特征提取阶段更加关注小目标的空间和梯度流信息; 在特征融合阶段, 提出一种多尺度跨层小目标特征融合网络(MCSTF-Net)来替换路径聚合网络(PANet), 在提高模型对PCB缺陷检测精度的同时大幅降低模型的参数量; 为进一步提升模型对小目标特征的理解程度, 将CCG网络和MCSTF-Net分别与SE(Squeeze-and-Excitation)注意力机制相结合, 在抑制无关通道的同时突出目标特征信息丰富的通道。消融实验结果表明, YOLOv5s-CMS模型对PCB缺陷检测的精确率、召回率、mAP@0.5和mAP@0.5∶0.95分别达到了98.1%、97.8%、98.4%和61.2%, 相比于YOLOv5s原模型分别提高了2.2、1.3、0.8和5.0百分点, 模型参数量同比减少了约46.1%。

关键词: 计算机视觉, YOLOv5s模型, 多尺度跨层特征融合, 印刷电路板缺陷, 小目标检测

Abstract:

The quality of Printed Circuit Boards (PCBs) plays a decisive role in the performance of industrial electronic products. Therefore, strict control of PCB factory quality is of considerable importance. The six types of defects that PCB are prone to are the primary basis for evaluating PCB quality. To address problems such as the low accuracy and large size of PCB defects detected by existing target detection algorithm models, an improved YOLOv5s-CMS model is proposed. According to the characteristics of PCB defects, the YOLOv5s-CMS model adopts a feature extraction network C2f-C3-Ghost (CCG) focusing on small target information to replace the original feature extraction network, such that the model pays more attention to the space and gradient flow information of small targets in the feature extraction stage. In the feature fusion phase, a Multi-scale Cross-layer Small Target Feature Fusion Network (MCSTF-Net) is proposed to replace the Path Aggregation Network (PANet), which can improve the accuracy of PCB defect detection while considerably reducing the number of parameters in the model. To further improve the model's understanding of small target characteristics, the CCG network and MCSTF-Net are combined with the Squeeze-and-Excitation (SE) attention mechanism to highlight the channels rich in target characteristics while suppressing irrelevant channels. The ablation experiment results showed that the accuracy, recall, mAP@0.5, and mAP@0.5∶0.95 of PCB defect detection using the YOLOv5s-CMS model reached 98.1%, 97.8%, 98.4%, and 61.2%, respectively. Compared to the original YOLOv5s model, the number of parameters increased by 2.2, 1.3, 0.8, and 5.0 percentage points, respectively, and the number of model parameters decreased by approximately 46.1%.

Key words: computer vision, YOLOv5s model, multi-scale cross-layer feature fusion, Printed Circuit Board (PCB) defect, small target detection

魏文泉, 莫宏伟. 基于改进YOLOv5s的PCB缺陷检测算法[J]. 计算机工程, 2026, 52(5): 226-238.

WEI Wenquan, MO Hongwei. PCB Defect Detection Algorithm Based on Improved YOLOv5s[J]. Computer Engineering, 2026, 52(5): 226-238.

https://www.ecice06.com/CN/Y2026/V52/I5/226

图/表 15

图1 YOLOv5s模型结构

Fig.1 YOLOv5s model structure

图2 改进后的YOLOv5s-CMS模型结构

Fig.2 Improved YOLOv5s-CMS model structure

图3 C2f模块结构

Fig.3 C2f module structure

图4 标准卷积与Ghost卷积

Fig.4 Standard convolution and Ghost convolution

图5 C3Ghost模块结构

Fig.5 C3Ghost module structure

图6 PANet和BiFPN结构图

Fig.6 PANet and BiFPN structure diagrams

图7 改进前后模型的特征融合网络架构对比

Fig.7 Comparison of feature fusion network architecture of the model before and after improvement

图8 SE注意力机制结构

Fig.8 SE attention mechanism structure

图9 PCB各类缺陷图

Fig.9 Various PCB defect diagrams

图10 数据增强后的部分样图

Fig.10 Partial sample images after data enhancement

图11 不同模型对PCB各类缺陷的检测效果

Fig.11 The detection effect of different models on various PCB defects

图12 训练阶段的YOLOv5s-CMS模型权重在测试集上的表现

Fig.12 Performance of YOLOv5s-CMS model weights on the test set during the training phase

图13 改进前后的模型在训练阶段的拟合度随训练轮数变化

Fig.13 The fit degree of the model before and after improvement changes with the number of training rounds

参考文献 26

1	张果, 陈逃, 王剑平, 等. 轻量化的PCB表面缺陷检测算法. 北京邮电大学学报, 2024, 47 (2): 38- 44.
	ZHANG G , CHEN T , WANG J P , et al. Lightweight PCB surface defect detection algorithm. Journal of Beijing University of Posts and Telecommunications, 2024, 47 (2): 38- 44.
2	DAI W T , MUJEEB A , ERDT M , et al. Soldering defect detection in automatic optical inspection. Advanced Engineering Informatics, 2020, 43, 101004.
3	GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Columbus, USA: IEEE Press, 2014: 580-587.
4	GIRSHICK R. Fast R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision. Santiago, Chile: IEEE Press, 2015: 1440-1448.
5	REN S Q , HE K M , GIRSHICK R , et al. Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39 (6): 1137- 1149. doi: 10.1109/TPAMI.2016.2577031
6	WEI L, DRAGOMIR A, DUMITRU E, et al. SSD: Single Shot MultiBox Detector[C]//Proceedings of ECCV 2016. Berlin, Germany: Springer, 2016: 21-37.
7	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. Las Vegas, USA: IEEE Press, 2016: 779-788.
8	BOCHKOVSKIY A, WANG C Y, LIAO H Y M. YOLOv4: optimal speed and accuracy of object detection[EB/OL]. [2024-09-14]. https://arxiv.org/abs/2004.10934.
9	LI C, LI L, JIANG H, et al. YOLOv6: a single-stage object detection framework for industrial applications[EB/OL]. [2024-09-14]. https://arxiv.org/abs/2209.02976.
10	ZHANG J, HUI L, LU J F, et al. Attention-based neural network for traffic sign detection[C]//Proceedings of the 24th International Conference on Pattern Recognition. Beijing, China: IEEE Press, 2018: 1839-1844.
11	陈仁祥, 詹赞, 胡小林, 等. 基于多注意力Faster RCNN的噪声干扰下印刷电路板缺陷检测. 仪器仪表学报, 2021, 42 (12): 167- 174.
	CHEN R X , ZHAN Z , HU X L , et al. Printed circuit board defect detection based on the multi-attentive Faster RCNN under noise interference. Chinese Journal of Scientific Instrument, 2021, 42 (12): 167- 174.
12	LI Y H , YANG F , LI Y , et al. Circuit breaker identification based on SSD. Journal of Physics: Conference Series, 2023, 2418 (1): 012080. doi: 10.1088/1742-6596/2418/1/012080
13	TANG J , WANG Z D , ZHANG H Y , et al. A lightweight surface defect detection framework combined with dual-domain attention mechanism. Expert Systems with Applications, 2024, 238, 121726. doi: 10.1016/j.eswa.2023.121726
14	周彦, 孟江南, 王冬丽, 等. 基于多尺度轻量化注意力的钢材缺陷检测. 控制与决策, 2024, 39 (3): 901- 909.
	ZHOU Y , MENG J N , WANG D L , et al. Steel defect detection based on multi-scale lightweight attention. Control and Decision, 2024, 39 (3): 901- 909.
15	ZENG N , WU P , WANG Z , et al. A small-sized object detection oriented multi-scale feature fusion approach with application to defect detection. IEEE Transactions on Instrumentation and Measurement, 2022, 71, 1- 14.
16	马淦, 谷雨, 彭冬亮. 结合改进YOLOv5s和动态数据增强的海面舰船检测. 计算机工程, 2025, 51 (9): 294- 305. doi: 10.19678/j.issn.1000-3428.0069459
	MA G , GU Y , PENG D L . Combining Improved YOLOv5s and dynamic data augmentation for sea surface ship detection. Computer Engineering, 2025, 51 (9): 294- 305. doi: 10.19678/j.issn.1000-3428.0069459
17	VARGHESE R, SAMBATH M. YOLOv8: a novel object detection algorithm with enhanced performance and robustness[C]//Proceedings of the International Conference on Advances in Data Engineering and Intelligent Computing Systems. Chennai, India: IEEE Press, 2024: 1-6.
18	HAN K, WANG Y H, TIAN Q, et al. GhostNet: more features from cheap operations[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle, USA: IEEE Press, 2020: 1577-1586.
19	LIU S, QI L, QIN H F, et al. Path aggregation network for instance segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE Press, 2018: 8759-8768.
20	HU J, SHEN L, SUN G. Squeeze-and-Excitation networks[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE Press, 2018: 7132-7141.
21	HOWARD A G, ZHU M, CHEN B, et al. MobileNets: efficient convolutional neural networks for mobile vision applications[EB/OL]. [2024-09-14]. https://arxiv.org/abs/1704.04861.
22	TAN M X, PANG R M, LE Q V. EfficientDet: scalable and efficient object detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle, USA: IEEE Press, 2020: 10778-10787.
23	LIN T Y, DOLLAR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, USA: IEEE Press, 2017: 936-944.
24	HUANG W B , WEI P . A PCB dataset for defects detection and classification. Journal of Latex Class Files, 2018, 14 (8): 1- 9.
25	REDMON J, FARHADI A. YOLOv3: an incremental improvement[EB/OL]. [2024-09-14]. https://arxiv.org/abs/1804.02767.
26	WANG C Y, BOCHKOVSKIY A, LIAO H M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver, Canada: IEEE Press, 2023: 7464-7475.

[1]	李沂杨, 陆声链, 王继杰, 陈明. 基于Transformer的DETR目标检测算法综述[J]. 计算机工程, 2026, 52(4): 62-81.
[2]	汤伟博, 方强, 李沛根, 艾龙金, 熊金红, 夏海廷. 基于RSD-YOLO的无人机航拍图像小目标检测[J]. 计算机工程, 2026, 52(4): 214-228.
[3]	曹继卫, 罗飞, 丁炜超. BS-YOLO: 基于BSAM注意力机制和SCConv的小目标检测算法[J]. 计算机工程, 2026, 52(3): 119-127.
[4]	唐克, 魏飞鸣, 李东瀛, 郁文贤. 基于改进YOLOv8的轻量化无人机图像目标检测算法[J]. 计算机工程, 2026, 52(3): 97-106.
[5]	张信佳, 王芳. 基于多层次特征融合和注意力机制的无人机图像小目标检测算法[J]. 计算机工程, 2026, 52(2): 148-157.
[6]	秦颖鑫, 张可佳, 潘海为, 巨亚昊. 计算机视觉对抗攻击研究综述[J]. 计算机工程, 2026, 52(2): 46-68.
[7]	王舒梦, 徐慧英, 朱信忠, 黄晓, 宋杰, 李毅. 基于改进YOLOv8n的航拍轻量化小目标检测算法: PECS-YOLO[J]. 计算机工程, 2025, 51(9): 280-293.
[8]	马淦, 谷雨, 彭冬亮. 结合改进YOLOv5s和动态数据增强的海面舰船检测[J]. 计算机工程, 2025, 51(9): 294-305.
[9]	倪源松, 韩军, 邹小燕, 胡广怡, 王文帅. 两阶段自适应分块输电线路螺栓缺陷检测方法[J]. 计算机工程, 2025, 51(8): 281-291.
[10]	栾孟娜, 郑秋梅, 王风华. 基于DMC-YOLO的交通标志实时检测算法[J]. 计算机工程, 2025, 51(7): 90-99.
[11]	奚琦, 王明杰, 魏敬和, 赵伟. 基于改进YOLOv3的航拍小目标检测算法[J]. 计算机工程, 2025, 51(6): 184-192.
[12]	黄昆, 齐肇建, 王娟敏, 胡倩, 胡伟超, 皮建勇. 基于改进YOLOv8的密集行人检测模型[J]. 计算机工程, 2025, 51(5): 133-142.
[13]	许华杰, 郑力文, 张品, 秦远卓. 基于多维注意力模块的轻量化混凝土裂缝检测方法[J]. 计算机工程, 2025, 51(5): 351-360.
[14]	周思瑜, 徐慧英, 朱信忠, 黄晓, 盛轲, 曹雨淇, 陈晨. 基于改进YOLOv8n的手机屏幕瑕疵检测算法: PGS-YOLO[J]. 计算机工程, 2025, 51(5): 326-339.
[15]	袁亚剑, 毛力. 一种增强前景的轻量级交通标志检测模型[J]. 计算机工程, 2025, 51(3): 54-63.

选择文件类型/文献管理软件名称

选择包含的内容