Improved YOLO-based image segmentation method for AEC welding cup profile

ZHANG Hongpu; LIU Xiyan; ZHAO Fengzhi; YAN Xiyan; FENG Yan

doi:10.12299/jsues.24-0125

Volume 39 Issue 3

Sep. 2025

Turn off MathJax

Article Contents

Article Navigation > Journal of Shanghai University of Engineering Science > 2025 > 39(3): 366-374

ZHANG Hongpu, LIU Xiyan, ZHAO Fengzhi, YAN Xiyan, FENG Yan. Improved YOLO-based image segmentation method for AEC welding cup profile[J]. Journal of Shanghai University of Engineering Science, 2025, 39(3): 366-374. doi: 10.12299/jsues.24-0125

Citation:

ZHANG Hongpu, LIU Xiyan, ZHAO Fengzhi, YAN Xiyan, FENG Yan. Improved YOLO-based image segmentation method for AEC welding cup profile[J]. Journal of Shanghai University of Engineering Science, 2025, 39(3): 366-374. doi: 10.12299/jsues.24-0125

Citation:

PDF( 2705 KB)

Improved YOLO-based image segmentation method for AEC welding cup profile

doi: 10.12299/jsues.24-0125

ZHANG Hongpu^{1, 2
,},
LIU Xiyan^{1, 2},
ZHAO Fengzhi^{1, 2},
YAN Xiyan^{1, 2},
FENG Yan^{1, 2
,
,}

1.
School of Mechanical and Automotive Engineering, Shanghai University of Engineering Science, Shanghai 201620, China
2.
Shanghai Large-Scale Component Intelligent Manufacturing Robot Technology Collaborative Innovation Center, Shanghai 201620, China

Received Date: 2024-04-29
Available Online: 2025-12-22
Publish Date: 2025-09-30

Abstract

Abstract

To achieve the accurate positioning of automated welding of aviation electrical connector (AEC), a method for the detection and segmentation of welding cup profiles was proposed based on machine learning. The effectiveness of feature extraction and prediction accuracy of the original network model were enhanced by incorporating a small target detection layer, the CBAM mechanism, and the GhostNet network. Concurrently, the number of parameters and space size of the improved model were reduced. The experimental results show that the improved YOLOv5s-Seg model achieves mean average precision of 84.2% and 44.6%. Compared with the original YOLOv5s model, this represents improved by 5.5% and 1.3%, respectively. The detection-segmentation method proposed effectively balances precision and speed, facilitating practical application and equipment deployment, and provides a theoretical basis for advancing the automated welding of AEC based on machine vision.
- machine learning,
- deep learning,
- aviation electrical connector (AEC),
- image segmentation

FullText(HTML)

References(17)

References

[1]	王腾, 郁大照, 朱蒙, 等. 海军某型飞机典型电连接器失效分析[J] . 装备环境工程, 2019, 16(12): 28 − 35.
[2]	杨奋为. 电连接器行业基础共性技术展望[J] . 机电元件, 2023, 43(6): 53 − 60,63. doi: 10.3969/j.issn.1000-6133.2023.06.015
[3]	牛步钊, 唐春燕, 杨存平, 等. 轨道交通装备用电连接器典型失效类型分析[J] . 智慧轨道交通, 2023, 60(3): 67 − 71. doi: 10.3969/j.issn.2097-0366.2023.03.013
[4]	童宣科. 基于图像处理的航空电连接器识别与焊点定位技术研究[D] . 桂林: 桂林电子科技大学, 2018.
[5]	孙逸鹤. 基于机器视觉电连接器自动焊接技术研究[D] . 桂林: 桂林电子科技大学, 2018.
[6]	单根立, 董沛森. 基于工业PC和PLC的圆形电连接器与多芯电缆自动焊接设备设计[J] . 机床与液压, 2020, 48(2): 94 − 97. doi: 10.3969/j.issn.1001-3881.2020.02.021
[7]	SONG L M, KANG J W, ZHANG Q L, et al. A weld feature points detection method based on improved YOLO for welding robots in strong noise environment[J] . Signal, Image and Video Processing, 2023, 17(5): 1801 − 1809. doi: 10.1007/s11760-022-02391-0
[8]	GAO A, FAN Z X, LI A N, et al. YOLO-Weld: a modified YOLOv5-based weld feature detection network for extreme weld noise[J] . Sensors, 2023, 23(12): 5640. doi: 10.3390/s23125640
[9]	LI M, HUANG J, XUE L, et al. A guidance system for robotic welding based on an improved YOLOv5 algorithm with a RealSense depth camera[J] . Scientific Reports, 2023, 13(1): 21299.
[10]	HAN K, WANG Y H, TIAN Q, et al. GhostNet: more features from cheap operations[C] //Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 1577 − 1586.
[11]	WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C] //Proceedings of the 15th European Conference on Computer Vision. Munich: Springer, 2018: 3 − 19.
[12]	YANG L X, ZHANG R Y, LI L D, et al. SimAM: a simple, parameter-free attention module for convolutional neural networks[C] //Proceedings of the 38th International Conference on Machine Learning. [S.l.] : PMLR, 2021: 11863 − 11874.
[13]	WANG Q L, WU B G, ZHU P F, et al. ECA-Net: efficient channel attention for deep convolutional neural networks[C] //Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 11531 − 11539.
[14]	HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C] //Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 7132 − 7141.
[15]	MA N N, ZHANG X Y, ZHENG H T, et al. ShuffleNet V2: practical guidelines for efficient CNN architecture design[C] //Proceedings of the 15th European Conference on Computer Vision. Munich: Springer, 2018: 122 − 138.
[16]	HOWARD A, SANDLER M, CHEN B, et al. Searching for MobileNetV3[C] //Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 1314 − 1324.
[17]	CHEN C P, GUO Z C, ZENG H E, et al. RepGhost: a hardware-efficient ghost module via reparameteri-zation[EB/OL] . (2022-11-11)[2024-01-31] . https://doi.org/10.48550/arXiv.2211.06088.