MIG weld seam tracking system based on image automatic enhancement and attention mechanism deep learning

被引：0

作者：

Zhu, Ming ^{[1
,2
]}

Lei, Runji ^{[1
]}

Weng, Jun ^{[1
]}

Wang, Jincheng ^{[1
]}

Shi, Yu ^{[1
,2
]}

机构：

[1] State Key Laboratory of Advanced Processing and Recycling of Non-ferrous Metals, Lanzhou University of Technology, Lanzhou

[2] Key Laboratory of Non-ferrous Metal Alloys and Processing of State Education Ministry, Lanzhou University of Technology, Lanzhou

来源：

Hanjie Xuebao/Transactions of the China Welding Institution | 2024年 / 45卷 / 11期

关键词：

deep learning; image enhancement; passive vision; weld tracking;

D O I：

10.12073/j.hjxb.20240718002

中图分类号：

学科分类号：

摘要：

Aiming at the problem that conventional MIG welding is difficult to adjust the welding position in real time according to the group deviation and thermal accumulation deformation, a weld seam tracking method based on passive vision is proposed. Through the image spatial domain filtering and automatic enhancement algorithm, the YOLO v7 deep learning model with attention mechanism is used to extract and analyze the groove alignment position and arc position in the region of interest in real time. The fuzzy control algorithm is used to control the MIG welding process in real time when the preset deviation occurs. The results show that, the image automatic enhancement algorithm is used to complete the preprocessing of the image, and the pixel gray value of the edge position is increased from 40 to about 110, which significantly improves the accuracy of the edge position information extraction; Based on the YOLO v7 network structure, the attention mechanism module is added to improve the efficiency of target detection, and the mAP index is as high as 99.27%. The preset deviation test shows that the pixel error of the alignment deviation detection is within 8 pixels, and the alignment deviation distance is controlled between ± 0.5 mm. © 2024 Harbin Research Institute of Welding. All rights reserved.

引用

页码：90 / 94

页数：4

共 7 条

[1] Han Qinglin, Li Dayong, Li Xinlei, Et al., Bead geometry measurement for wire and arc additive manufacturing using active-passive composite vision sensing based on regional filter, Transactions of the China Welding Institution, 41, 9, pp. 28-32, (2020)
[2] Chen Huabin, Chen Shanben, Key information perception and control strategy of intellignet welding under complex scene, Acta Metallurgica Sinica, 58, 4, pp. 541-550, (2022)
[3] Xia Lei, Zhou Jianping, Xue Ruilei, Et al., Real-time seam tracking during narrow gap GMAW process based on the wide dynamic vision sensing method, Journal of Manufacturing Processes, 101, pp. 820-834, (2023)
[4] Xiao Runquan, Xu Yanling, Hou Zhen, Et al., A feature extraction algorithm based on improved Snake model for multi-pass seam tracking in robotic arc welding[J], Journal of Manufacturing Processes, 72, pp. 48-60, (2021)
[5] Zhang Guangjun, Len Xiaoyu, Wu Lin, Robotic welding seam tracing system based on structure light vision sensor, Transactions of the China Welding Institution, 29, 9, pp. 8-10, (2008)
[6] Wang Weixi, Satoshi Yamane, Wang Qi, Et al., Visual sensing and quality control in plasma MIG welding, Journal of Manufacturing Processes, 86, pp. 163-176, (2023)
[7] Zhang Zhifen, Wen Guangrui, Chen Shanben, Weld image deep learning-based on-line defects detection using convolutional neural networks for Al alloy in robotic arc welding[J], Journal of Manufacturing Processes, 45, pp. 208-216, (2019)

← 1 →