Abstract：Aiming at the low detection accuracy of IC pin soldering defects due to small target size and dense pins, an algorithm for IC pin soldering defects detection based on improved PP-YOLOv2 is proposed. By introducing SE attention mechanism behind the backbone network, the importance of different channels in the feature map is distinguished, the key features of the target area are strengthened, and the network feature extraction ability is improved. k-means++ clustering algorithm is used to generate 9 cluster centers to reduce the error impact on the detection results caused by improper random selection of initial cluster centers. The experimental results show that the average accuracy of the improved algorithm for detecting the defects of IC pin soldering short circuit, missing pin, warping pin and little tin is 97.9%, 96.1%, 96.7% and 95.8% respectively. Under the threshold value of 0.5, the average accuracy reaches 96.6%, which is 14.9% and 5.1% higher than YOLOv3 and PP-YOLOv2 respectively. The detection time of the improved algorithm for a single picture is 0.151s, which meets the speed requirements of IC quality inspection and provides a reference value for IC pin soldering defect detection.
李娜,王学影,胡晓峰,郭斌,罗哉. 基于改进PP-YOLOv2的IC引脚焊接缺陷检测算法研究[J]. 计量学报, 2023, 44(10): 1574-1581.
LI Na,WANG Xue-ying,HU Xiao-feng,GUO Bin,LUO Zai. Research on IC Pin soldering Defect Detection Algorithm Based on Improved PP-YOLOv2. Acta Metrologica Sinica, 2023, 44(10): 1574-1581.
Zhai J B, Chen J, Qin Z F. Common Problems and Corrective Measures in Factory Inspection of Electronic Product Printing Process ［J］. Quality and Certification, 2022, 6(6): 73-75.
Cai X B, Zhao Y, Tang Y Q. On Testing of PCB Test ［J］. Journal of Liaoning University of Technology, 2004, 24(4): 32-34.
Tang B, Kong J Y, Wu S Q. Review of Surface Defect Detection based on Machine Vision ［J］. China Journal of Image Graphics, 2017, 22(12): 1640-1663.
Malge P S, Nadaf R S. PCB defect detection, classification and localization using mathematical morphology and image processing tools ［J］. International journal of computer applications, 2014, 87(9): 40-45.
［5］
Kumar M, Singh N K, Kumar M, et al. A novel approach of standard data base generation for defect detection in bare PCB ［C］//International Conference on Computing, Communication & Automation. IEEE, Greater Noida, India, 2015: 11-15.
Girshick R. Fast r-cnn ［C］//Proceedings of the IEEE International Conference on Computer Vision, Santiago, Chile, 2015: 1440-1448.
［24］
Misra D. Mish: A self regularized non-monotonic neural activation function ［DB］.https://arxiv. org/abs/1908.08681, 2019.
［6］
Gaidhane V H, Hote Y V, Singh V. An efficient similarity measure approach for PCB surface defect detection ［J］. Pattern Analysis and Applications, 2018, 21(1): 277-289.
Tian Z, She Q S, Gan H T, et al. DCNN for Task Coordination of Facial Landmark Localization and Head Pose Estimation ［J］. Acta Metrologica Sinica, 2019, 40(4): 576-582.
Ding R, Dai L, Li G, et al. TDD-net: a tiny defect detection network for printed circuit boards ［J］. CAAI Transactions on Intelligence Technology, 2019, 4(2): 110-116.
Li C F, Cai J L, Qiu S H, et al. Defect detection of PCB based on improved YOLOv4 algorithm ［J］. Electronic measurement technology, 2021, 44(17): 146-153.
［15］
Ren S, He K, Girshick R, et al. Faster r-cnn: Towards real-time object detection with region proposal networks ［J］. Advances in Neural Information Processing Systems, 2015, 28(6): 1137-1149.
［17］
Redmon J, Farhadi A. Yolov3: An incremental improvement ［DB］.https://arxiv. org/abs/ 1908.08681, 2019.
［19］
Huang X, Wang X, Lv W, et al. PP-YOLOv2: A practical object detector ［DB］. https://arxiv.org/abs/2104.10419v1, 2021.
［21］
He T, Zhang Z, Zhang H, et al. Bag of tricks for image classification with convolutional neural networks ［C］//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, USA, 2019: 558-567.
［22］
Liu S, Qi L, Qin H, et al. Path aggregation network for instance segmentation ［C］//Proceedings of the IEEE conference on computer vision and pattern recognition, Salt Lake City, USA, 2018: 8759-8768.
Wang Z W, Guo B, Hu X F, et al. Internal Groove Defect Detection Method of Brake Master Cylinder Based on FCOS Neural Network ［J］. Acta Metrologica Sinica, 2021, 42(9): 1225-1231.
Chen J Q, Jin X H, Wang W Y, et al. Vehicle Flow Detection Based on YOLOv3 and Deep Sort［J］. Acta Metrologica Sinica, 2021, 42(6): 718-723.
［13］
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, 2014: 580-587.
［16］
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, 2016: 779-788.
［18］
Liu W, Anguelov D, Erhan D, et al. Ssd: Single shot multibox detector ［C］//European Conference on Computer Vision, Amsterdam, Netherlands, 2016: 21-37.
［20］
He K, Zhang X, Ren S, et al. Deep residual learning for image recognition ［C］//Proceedings of the IEEE conference on computer vision and pattern recognition, Las Vegas, USA, 2016: 770-778.
［23］
Hu J, Shen L, Sun G. Squeeze-and-excitation networks ［C］//Proceedings of the IEEE conference on computer vision and pattern recognition, Salt Lake City, USA, 2018: 7132-7141.
2023, Vol. 44 
