Abstract
Welding is an important joining technology but the defects in welds wreck the quality of the product evidently. Weld defect detection is always an important and challenging research due to the various defects with complex background. A modified lightweight U-Net with attention mechanism model (LWAU-Net) is constructed for weld defect detection. In the model, the multiple convolutional and pooling kernels with different sizes are utilized to learn the multi-scale discriminant features, and the attention mechanism is used to capture to adaptively select multi-scale features for classification. Compared with the standard convolution neural networks (CNN), LWAU-Net is integrated to learn the multi-scale features for weld defect detection, especially small defects. Experiment results on the weld defect image dataset show that the proposed method outperforms the state-of-the-art method on the same dataset and the obtained multi-size defect edge is clearer.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Gang, T., Takahashi, Y., Wu, L.: Intelligent pattern recognition and diagnosis of ultrasonic inspection of welding defects based on neural network and information fusion. Sci. Technol. Weld. Joining 7(5), 314–320 (2002)
Dinham, M., Gu, F.: Autonomous weld seam identification and localization using eye-in-hand stereo vision for robotic arc welding. Robot. Comput.-Integr. Manuf. 29(5), 288–301 (2013)
Zhang, L., Ye, Q., Yang, W., et al.: Weld Line Detection and Tracking via Spatial. Temporal Cascaded Hidden Markov Models and Cross Structured Light. IEEE Trans. Instr. Measur. 63(4), 742–753 (2014)
He, Y.: Weld seam profile detection and feature point extraction for multi-pass route planning based on visual attention model. Robot. Comput. Integ. Manuf. 37, 251–261 (2015)
Leemans, V., Destain, M.F.: Line cluster detection using a variant of the Hough transform for culture row localization. Image Vis. Comput. 24(5), 541–550 (2016)
Shao, W.J., Huang, Y., Zhang, Y.: A novel weld seam detection method for space weld seam of narrow butt joint in laser welding. Optics Laser Technol. 99, 39–51 (2018)
Akram, J., Kalvala, P.R., Chalavadi, P., Misra, M.: Dissimilar metal weld joints of P91/Ni alloy: microstructural characterization of HAZ of P91 and stress analysis at the weld interfaces. J. Mater. Eng. Perform. 27(8), 4115–4128 (2018). https://doi.org/10.1007/s11665-018-3502-8
Kulkarni, A., Dwivedi, D.K., Vasudevan, M.: Study of mechanism, microstructure and mechanical properties of activated flux TIG welded P91 Steel-P22 steel dissimilar metal joint. Mater. Sci. Eng. A. 731, 309–323 (2018)
Sirohi, S., Kumar, S., Bhanu, V., et al.: Study on the Variation in Mechanical Properties along the Dissimilar Weldments of P22 and P91 Steel. J. Mater. Eng. Perform. 31, 2281–2296 (2022)
Mu, W., Gao, J., Jiang, H., et al.: Automatic classification approach to weld defects based on PCA and SVM. Insight Non Destructive Testing & Condition Monitoring 55(10), 535–539 (2013)
Mu, W., Liu, G., Peng, L., et al.: A novel classification approach of weld defects based on dual-parameters optimization of PCA and LDA. In: International Conference on Advances in Mechanical Engineering & Industrial Informatics, pp. 1425–1429 (2015)
Murugan, R., Venugobal, P.R., Ramaswami, T.P., et al.: Studies on the effect of weld defect on the fatigue behavior of welded structures. China Weld. 27(1), 53–59 (2018)
Cheng, Y., Wang, Q., Jiao, W., et al.: Detecting dynamic development of weld pool using machine learning from innovative composite images for adaptive welding. J. Manuf. Process. 56, 908–915 (2020)
Yazid, H., Arof, H., Yazid, H., et al.: Weld detect identification using texture features and dynamic time warping. Appl. Mech. Mater. 752–753, 1045–1050 (2015)
Wei, A., Chang, B., Xue, B., et al.: Research on the weld position detection method for sandwich structures from face-panel side based on backscattered X-ray. Sensors 19(14), 3198 (2019)
Zhang, P.L., Zhao, Z.Q., Wang, Y.P.: X-Ray testing of weld defect of automatic recognition and alarm technology research. Appl. Mech. Mater. 472, 495–502 (2014)
Ren, J., Wang, Y.: Overview of object detection algorithms using convolutional neural networks. J. Comput. Commun. 10, 115–132 (2022)
Khumaidi, A., Yuniarno, E.M., Purnomo, M.H.: Welding defect classification based on convolution neural network (CNN) and Gaussian kernel. In: International Seminar on Intelligent Technology and Its Applications (ISITIA), pp. 261–265 (2017)
Zhang, H., Chen, Z., Zhang, C., et al.: Weld defect detection based on deep learning method. In: IEEE 15th International Conference on Automation Science and Engineering (CASE), pp. 1574–1579 (2019)
Zhang, L.F., Gao, W.X., Wang, Z., et al.: Research on weld defect identification with X-ray based on convolutional neural network. J. Phys: Conf. Ser. 1894(1), 012071 (2021)
Liu, M., Xie, J., Hao, J., et al.: A lightweight and accurate recognition framework for signs of X-ray weld images. Comput Ind 135, 103559 (2021)
Jiang, H., Hu, Q., Zhi, Z., et al.: Convolution neural network model with improved pooling strategy and feature selection for weld defect recognition. Welding in the World, Le Soudage Dans Le Monde, 65(4), pp. 731–744 (2020)
Acknowledgments
This work is supported by the National Natural Science Foundation of China (Nos. 62172338 and 62072378).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Huang, L. et al. (2022). Modified Lightweight U-Net with Attention Mechanism for Weld Defect Detection. In: Huang, DS., Jo, KH., Jing, J., Premaratne, P., Bevilacqua, V., Hussain, A. (eds) Intelligent Computing Theories and Application. ICIC 2022. Lecture Notes in Computer Science, vol 13393. Springer, Cham. https://doi.org/10.1007/978-3-031-13870-6_25
Download citation
DOI: https://doi.org/10.1007/978-3-031-13870-6_25
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-13869-0
Online ISBN: 978-3-031-13870-6
eBook Packages: Computer ScienceComputer Science (R0)