Citation: | Li Guodong, Wu Zhisheng, Peng Furong, et al. X-ray detection method of weld defects based on supervised contrastive learning[J]. Welding & Joining, 2024(7):7 − 14. DOI: 10.12073/j.hj.20230908002 |
[Objective] Deep learning-based surface defect detection algorithms are widely used in surface defect detection. However, in the field of weld defect detection, appearance characteristics of weld defects are characterized by a large deviation of samples of the same category and a small deviation of samples of different categories, which poses a challenge to the effective recognition of weld defects. [Methods] To this end, a supervised contrastive learning method for weld defect detection (SCL-DD) was proposed to extend supervised contrastive learning to the field of weld defect detection, where effective similarity computation was carried out through positive and negative samples, so that defect samples of the same class were closer to each other in the embedding space, and defects from different classes were far away from each other, and negative effects of the interclass bias and cross-class bias on detection performance were reduced. [Results] A cosine classifier was introduced to improve detection performance of differential defect samples by calculating the cosine similarity between feature encoding and classification prototype. Performance of the proposed method was validated on a steel pipe weld surface defect dataset. [Conclusion] SCL-DD method had an average accuracy of 96.9, which was better than other advanced deep learning networks.
[1] |
张龙飞, 高炜欣, 冯小星. 基于卷积神经网络X射线环焊缝缺陷检测[J]. 焊接, 2022(3): 26 − 34.
Zhang Longfei, Gao Weixin, Feng Xiaoxing. X-ray circumferentialweld defect detection based on convolutional neural network [J]. Welding & Joining, 2022(3): 26 − 34.
|
[2] |
Li Dawei, Li Yida, Xie Qian, et al. Tiny defect detection in high-resolution aero-engine blade images via a coarse-to-fine framework[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1 − 12.
|
[3] |
Liu W, Anguelov D, Erhan D, et al. SSD: single shot multibox detector [C]//Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, 2016: 21 − 37.
|
[4] |
Chen Qiang, Wang Yingming, Yang Tong, et al. You only look one-level feature [C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 2021: 13039 − 13048.
|
[5] |
Bhattacharya G, Mandal B, Puhan N B. Interleaved deep artifacts-aware attention mechanism for concrete structural defect classification[J]. IEEE Transactions on Image Processing, 2021, 30: 6957 − 6969. doi: 10.1109/TIP.2021.3100556
|
[6] |
Tu Zhenwei, Wu Songrong, Kang Gaoqiang, et al. Real-time defect detection of track components: considering class imbalance and subtle difference between classes[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1 − 12.
|
[7] |
Gidaris S, Komodakis N. Dynamic few-shot visual learning without forgetting[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 2018: 4367 − 4375.
|
[8] |
Qi H, Brown M, Lowe D G. Low-shot learning with imprinted weights[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 2018: 5822 − 5830.
|
[9] |
鲍峰, 王俊红, 张锋, 等. 基于YOLO V3的管道环焊缝缺陷检测[J]. 焊接, 2021(8): 56 − 61.
Bao Feng, Wang Junhong, Zhang Feng, et al. Pipeline girth weld defect detection and recognition based on YOLO V3[J]. Welding & Joining, 2021(8): 56 − 61.
|
[10] |
赵新玉, 马小创, 李正光, 等. 焊接缺陷识别中的多模态注意力方法[J]. 焊接, 2022(12): 49 − 54.
Zhao Xinyu, Ma Xiaochuang, Li Zhengguang, et al. Multi-modal attention method in welding defect recognition[J]. Welding & Joining, 2022(12): 49 − 54.
|
[11] |
王睿, 胡云雷, 刘卫朋, 等. 基于边缘AI的焊缝X射线图像缺陷检测[J]. 焊接学报, 2022, 43(1): 79 − 84.
Wang Rui, Hu Yunlei, Liu Weipeng, et al. Defect detection of weld X-ray image based on edge AI[J]. Transactions of the China Welding Institution, 2022, 43(1): 79 − 84.
|
[12] |
Chen T, Kornblith S, Norouzi M, et al. A simple framework for contrastive learning of visual representations[J]. Proceedings of the 37th International Conference on Machine Learning, PMLR, 2020, 119: 1597 − 1607.
|
[13] |
He Kaiming, Fan Haoqi, Wu Yuxin, et al. Momentum contrast for unsupervised visual representation learning [C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 2020: 9729 − 9738.
|
[14] |
Khosla P, Teterwak P, Wang C, et al. Supervised contrastive learning[J]. Advances in neural information processing systems, 2020, 33: 18661 − 18673.
|
[15] |
Bukchin G, Schwartz E, Saenko K, et al. Fine-grained angular contrastive learning with coarse labels [C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 2021: 8730 − 8740.
|
[16] |
Lo Y C, Chang C C, Chiu H C, et al. ClCC: contrastive learning for color constancy [C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 2021: 8053 − 8063.
|
[17] |
Liu Jinlu, Song Liang, Qin Yongqiang. Prototype rectification for few-shot learning [C]//Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, 2020: 741 − 756.
|
[18] |
Ren S Q, He K M, Girshick R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137 − 1149. doi: 10.1109/TPAMI.2016.2577031
|
[19] |
Sun Bo, Li Banghuai, Cai Shengcai, et al. FSCE: few-shot object detection via contrastive proposal encoding [C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 2021: 7352 − 7362.
|
[20] |
Feng Chengjian, Zhong Yujie, Gao Yu, et al. TOOD: task-aligned one-stage object detection [C]//2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada, 2021: 3490 − 3499.
|
[21] |
Dai Xiyang, Chen Yinpeng, Yang Jianwei, et al. Dynamic DETR: end-to-end object detection with dynamic attention [C]//2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada, 2021: 2988 − 2997.
|
[1] | Hu Wei, Ruan Xianhu, Jin Mingdong, Liu Duo, Zhang Jiandong. Weld defect recognition in ultrasonic TOFD image based on depth learning algorithm[J]. WELDING & JOINING, 2024, (10): 55-60. DOI: 10.12073/j.hj.20240228005 |
[2] | Ren Wenjian, Wang Yonghong, Li Chunkai, Shi Yu, Sun Zhongcheng, Liu Guozeng. Automatic detection method of DR digital imaging for brazing seam defects of thrust chamber body[J]. WELDING & JOINING, 2023, (7): 34-39, 53. DOI: 10.12073/j.hj.20220419002 |
[3] | Zhang Longfei, Gao Weixin, Feng Xiaoxing. X-ray circumferential weld defect detection based on convolutional neural network[J]. WELDING & JOINING, 2022, (3). DOI: 10.12073/j.hj.20211110003 |
[4] | Li Liangliang, Zheng Shiwei, Shan Qingqun, Zuo Yuda. Ultrasonic phased array detection and 3D visualization of deep defects in welded structural parts of bogies[J]. WELDING & JOINING, 2021, (9): 44-50. DOI: 10.12073/j.hj.20210705002 |
[5] | Fang Haiji, Ye Guoliang, Lü Bo, Zhang Yanxi, Gao Xiangdong. Laser ultrasonic detection for surface defect of stainless steel additive manufacturing parts[J]. WELDING & JOINING, 2021, (9): 16-21. DOI: 10.12073/j.hj.20210602001 |
[6] | SHOU Mingyao, YU Huanwei, ZHU Yuan, YU Xiang. Discussion on key factors of pressure vessel welding intelligence[J]. WELDING & JOINING, 2020, (11): 50-55. DOI: 10.12073/j.hj.20200703001 |
[7] | LI Mingchao, LI Guangquan, ZHOU Canfeng, CHEN Zhi, ZHOU Zhenzhen. Horizontal welding deformation of API X65 pipe in deep water J-laying[J]. WELDING & JOINING, 2020, (6): 12-17. DOI: 10.12073/j.hj.20200518004 |
[8] | CHI Dazhao, MA Ziqi, CHENG Yi, WANG Ziming, TANG Ziheng. Defect detection method for a single lap joint based on X-ray image processing[J]. WELDING & JOINING, 2019, (8): 1-4. DOI: 10.12073/j.hj.20190514002 |
[9] | SONG Fan, MA Jilong, YU Kang, ZHOU Jie. A study on porosity characteristics of laser deep penetration welding in space engines[J]. WELDING & JOINING, 2018, (8): 44-49. |
[10] | Wang Guoan1, Dai Jianshu2, Huang Haibo1, Liu Xiaohui3, Hou Guoqing1. Exploration on training mode of welding automation talents in intelligent manufacturing environment[J]. WELDING & JOINING, 2017, (7): 70-72. |