Abstract
When convolutional neural network (CNN) is used for welding defect detection image recognition, the recognition result will be affected by many factors such as human factors, the activation function is sensitive to input parameters, and the edge features are weakened. In order to overcome the above problems, the methods include image processing, exponential linear unit (ELU) activation function and improved pooling model are used. According to the experiment, the image processing method can effectively segment the weld and defects, and the defect location in the weld image can be located. Using the ELU activation function in the CNN model can improve the robustness of the neural network to the input parameters and increase the sparsity of the network to increase the model’s convergence speed. The improved pooling method based on grayscale adaptation can increase the extraction range of weld defect features and reduce the impact of noise, and has certain dynamic adaptability to the defect features. The result shows that the improved convolutional neural network(ICNN) method can effectively improve the accuracy of recognition in weld image recognition, and the overall recognition rate can reach 98.13%.
Similar content being viewed by others
References
Ali A, Zhu Y, Zakarya M (2021) A data aggregation based approach to exploit dynamic spatio-temporal correlations for citywide crowd flows prediction in fog computing [J]. Multimed Tools Appl 2:4–37
Bharodiya AK, Gonsai AM (2019) An improved edge detection algorithm for X-Ray images based on the statistical range. Heliyon 5:27–36
Ding F, An-de H, Jian-kang H et al (2020) X-ray image defect recognition method for pipe weld based on improved convolutional neural network [J]. Transactions of the China Weld Institution 41(01):7–11 +97
Federer C, Xu H, Fyshe A et al (2020) Improved object recognition using neural networks trained to mimic the brain's statistical properties - ScienceDirect [J]. Neural Netw 131:103–114
Gao L, Liao YL, Zhu BH et al (2007) Adaptive thresholds edge detection for image based on segmentation. Journal of Transport Information and Safety 25(5):73–76
Gorban AN, Mirkes EM, Tukin IY (2020) How deep should be the depth of convolutional neural networks: a backyard dog case study [J]. Cogn Comput 12(1):388–397
Gu J, Wang Z, Kuen J, Ma L, Shahroudy A, Shuai B, Liu T, Wang X, Wang G, Cai J, Chen T (2018) Recent advances in convolutional neural networks. Comput Sci 77:354–377
He K, Zhang X, Ren S et al (2014) Spatial pyramid pooling in deep convolutional networks for visual recognition. IEEE Transactions on Pattern Analysis & Machine Intelligence 37(9):1904–1920
Lecun Y, Bengio Y (2015) Hinton G. Deep Learning 521(7553):436–445
LeCun Y, Bottou L, Bengio Y et al (1998) Gradient-based learning applied to document recognition. Proc IEEE 86(11):2278–2324
Li T, Meng Z, Ni B, Shen J, Wang M (2016) Robust geometric ℓp-norm feature pooling for image classification and action recognition. Image Vis Comput 55(2):64–76
Li Y, Xie W, Li H (2017) Hyperspectral image reconstruction by deep convolutional neural network for classification. Pattern Recogn 63:371–383
Lin JY, Ke HR, Chi BC et al (2007) Designing a classifier by a layered multi-population genetic programming approach. Pattern Recogn 40(8):2211–2225
McIlhagga W (2010) The canny edge detector revisited. Int J Comput Vis 91(3):251–261
Mery D, Riffo V, Zscherpel U, Mondragón G, Lillo I, Zuccar I, Lobel H, Carrasco M (2015) GDXray: the database of X-ray images for nondestructive testing. J Nondestruct Eval 34(4):1–12
O'Shea TJ, Hoydis J (2017) An introduction to deep learning for the physical layer. IEEE Transactions on Cognitive Communications & Networking 3(4):563–575
Sakamoto M, Hiasa Y, Otake Y, Takao M, Suzuki Y, Sugano N, Sato Y (2020) Bayesian segmentation of hip and thigh muscles in metal artifact-contaminated CT using convolutional neural network-enhanced normalized metal artifact reduction [J]. Journal of VLSI Signal Processing Systems for Signal, Image, and Video Technology 92(3):335–344
SHAO W, LIU X, WU Z. (2019) A robust weld seam detection method based on particle filter for laser welding by using a passive vision sensor [J]. Int J Adv Manuf Technol 4:1–10
Sun Y, Song HH, Bai P et al (2004) Real-time automatic detection of weld defects in X-ray images. Transactions of the China Welding Institution 25(2):115–118
Wang F, Zhu JJ (2017) Research of machine vision for welding defect in complex environment. Welding Technology 46(5):127–133
Wei Y, Xia W, Lin M et al (2015) HCP: a flexible CNN framework for multi-label image classification. IEEE Trans Softw Eng 38(9):1901–1907
Wu X, He R, Sun ZN et al (2017) A light CNN for deep face representation with Noisy labels. IEEE Transactions on Information Forensics & Security 14(8):1–13
Yang Z, Kuang N, Fan L et al (2018) Review of image classification algorithms based on convolutional neural networks. J Signal Process 34(12):84–99
Yang M, Huang F, Lv X (2019) A feature learning approach for face recognition with robustness to noisy label based on top-N prediction. Neurocomputing 330(22):48–55
Yang L, Chen W, Liu W, Zha B, Zhu L (2020) Random noise attenuation based on residual convolutional neural network in seismic datasets [J]. IEEE Access 8:30271–30286
Yong-wei YU, Liu-qing DU, Zhe Y et al (2016) Dynamic detection of casting defects radiographic image based on deep learning feature. Transactions of the Chinese Society for Agricultural Machinery 47(7):407–412
Zhang ZT, Cheng DD, Wang JJ et al (2019) Quantum based subgraph convolutional neural networks. Pattern Recogn 88:38–49
Zhang Z, Ding S, Jia W (2019) A hybrid optimization algorithm based on cuckoo search and differential evolution for solving constrained engineering problems. Eng Appl Artif Intell 85:254–268
Zhang Z, Ding S, Sun Y (2020) A support vector regression model hybridized with chaotic krill herd algorithm and empirical mode decomposition for regression task. Neurocomputing 410:185–201
Zhou FY, Jin LP, Dong J (2017) Review of convolutional neural network. Chinese Journal of Computers 40(6):1229–1251
Acknowledgments
This work is funded by the National Natural Science Foundation of China (No. 01020607).
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Ethical approval
Not applicable.
Consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Hu, A., Wu, L., Huang, J. et al. Recognition of weld defects from X-ray images based on improved convolutional neural network. Multimed Tools Appl 81, 15085–15102 (2022). https://doi.org/10.1007/s11042-022-12546-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-12546-3