ABSTRACT
The1 spot welding technique is widely used in the industrial production line, but it suffers inconsistent quality. Therefore, the evaluation of the spot-welding product is of great importance for the industrial production. Many destructive and nondestructive methods have been used in the product evaluation, but they are inefficient and hard to be applied in the mass production. In recent year, machine vision method has been used to differentiate the acceptable and failed spot welding products according to their solder joint images. This opened new opportunities for the spot welding product quality evaluation using digital image technique. However, this method cannot achieve general performance on different spot-welding products as well as ideal classification accuracy. In this work, a novel method which based on the transfer learning technique was proposed to classify the spot-welding products according to their solder joint images. The GoogLeNet was used to extract the features of the solder joint image, which is pretrained on the ImageNet. Then a multilayer perceptron (MLP) was used to classify these images. Our method achieved a final classification accuracy of 96.99% on a testing set included 334 images.
- Ishaya ZD, Dauda M, and Pam GY, et al. 2013. Destructive testing and production system integrity. Advances in Applied Science Research, 4, 4(2013), 184--189.Google Scholar
- Xiaodong Wan, Yuanxun Wang and Dawei Zhao. 2018. Quality Estimation in Small Scale Resistance Spot Welding of Titanium Alloy Based on Dynamic Electrical Signals. ISIJ International, 58, 4(2018), 721--726.Google Scholar
Cross Ref
- Xingjue Wang, and Junhong Zhou, et al. 2018. Quality monitoring of spot welding with advanced signal processing and data-driven techniques. Transactions of the Institute of Measurement and Control, 40, 7(2018), 2291--2302.Google Scholar
Cross Ref
- Hongjie Zhang, and Yanyan Hou, et al. 2015. A new method for nondestructive quality evaluation of the resistance spot welding based on the radar chart method and the decision tree classifier. Int J Adv Manuf Technol, 78, 5--8(2015), 841--851.Google Scholar
- Samo Simončič and Primož Podržaj. 2012. Image-based electrode tip displacement in resistance spot welding. Measurement Science and Technology, 23, 6(2012).Google Scholar
- Marco Leo a, and Marco Del Coco a, et al. 2018. Automatic visual monitoring of welding procedure in stainless steel kegs. Optics and Lasers in Engineering, 104, SI(2018), 220--231.Google Scholar
- Wenhui Hou, and Ye Wei, et al. 2018. Automatic Detection of Welding Defects using Deep Neural Network. 10th International Conference on Computer and Electrical Engineering, Univ Alberta, Edmonton, CANADA.Google Scholar
- Zhuang Ni, and Yan Yan, et al. 2018. Multi-label learning based deep transfer neural network for facial attribute classification. Pattern Recognition, 80, 225--240.Google Scholar
Cross Ref
- Boufenar, Chaouki, Kerboua, Adlen, Batouche, and Mohamed. 2017. Investigation on deep learning for off-line handwritten Arabic character recognition. Cognitive Systems Research, 50, 180--195.Google Scholar
Cross Ref
- Lei, Haijun, and Han, Tao, et al. 2018. A deeply supervised residual network for HEp-2 cell classification via cross-modal transfer learning. Pattern Recognition, 79, 290--302.Google Scholar
Cross Ref
- G Christian Szegedy, and Wei Liu, et al. 2015. going deeper with convolutions. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA, 1--9.Google Scholar
Cross Ref
- Mohammed Al-Qizwini, and Iman Barjasteh, et al. 2017. Deep Learning Algorithm for Autonomous Driving using GoogLeNet. 2017 IEEE Intelligent Vehicles Symposium (IV), Redondo Beach, CA, 89--96.Google Scholar
Cross Ref
- Singla A, Yuan L, and Ebrahimi T. 2016. Food/Non-food Image Classification and Food Categorization using Pre-Trained GoogLeNet Model. International Workshop on Multimedia Assisted Dietary Management, ACM, Amsterdam, NETHERLANDS, 3--11. Google Scholar
Digital Library
- Zhuoyao Zhong, Lianwen Jin, and Zecheng Xie. 2015. High Performance Offline Handwritten Chinese Character Recognition Using GoogLeNet and Directional Feature Maps. 13th IAPR International Conference on Document Analysis and Recognition (ICDAR), Nancy, FRANCE, 846--850. Google Scholar
Digital Library
- Atkinson, P.M., and Tatnall, A.R.L. 1997. Introduction neural networks in remote sensing. International Journal of Remote Sensing, 18, 4(1997), 699--709.Google Scholar
Cross Ref
- F. Del Frate, F. Pacifici, G. Schiavon, and C. Solimini. 2007. Use of neural networks for automatic classification from high-resolution images. IEEE Trans. Geosci. Remote Sens, 45, 800--809.Google Scholar
Cross Ref
- Glorot, X., Bordes, A., and Bengio, Y. 2011. Deep sparse rectifier neural networks. International Conference on Artificial Intelligence and Statistics 2011, AISTATS, 15, 315--323.Google Scholar
- Wang, Feng, and Cheng, Jian, et al. 2018. Additive Margin Softmax for Face Verification. IEEE Signal Processing Letters, 25, 7(2018), 926--930.Google Scholar
Cross Ref
- SHORE, JE, and JOHNSON, RW. 1980. axiomatic derivation of the principle of maximum-entropy and the principle of minimum cross-entropy. IEEE Transactions on Information Theory, 26, 1(1980), 26--37. Google Scholar
Digital Library
- Hu, Kai, and Zhang, Zhenzhen, et al. 2018. Retinal vessel segmentation of color fundus images using multiscale convolutional neural network with an improved cross-entropy loss function. Nurocomputing, 309, 179--191.Google Scholar
Cross Ref
- Pereira, Sergio, Pinto, and Adriano, et al. 2016. Brain Tumor Segmentation Using Convolutional Neural Networks in MRI Images. IEEE Transactions on Medical Imaging, 35, 5(2016), 1240--1251.Google Scholar
Cross Ref
Index Terms
- An Evaluation Method of Acceptable and Failed Spot Welding Products Based on Image Classification with Transfer Learning Technique
Recommendations
Influence of Variation of Welding Parameters on Spot Welding Quality
ISIC '11: Proceedings of the 2011 International Conference on Information Security and Intelligence ControlU-I images of mild steel welding spot have ellipse, double-ended and multi-ended models. When no splash happens, the U-I images shapes ellipse. If one splash happens in the welding, U-I curves shapes double-ended. If many, multi-ended shapes were found. ...
Analysis of Welding Defects in Spot Welding Process U-I Curves
WGEC '09: Proceedings of the 2009 Third International Conference on Genetic and Evolutionary ComputingHigh speed collecting and managing system of spot welding signal can achieve the A/D conversion and data collection. The data of welding current, electrode voltage as well as welding cycle has been collected. Subsequently, they were processed with data ...
Pollen Grain Microscopic Image Classification Using an Ensemble of Fine-Tuned Deep Convolutional Neural Networks
Pattern Recognition. ICPR International Workshops and ChallengesAbstractPollen grain micrograph classification has multiple applications in medicine and biology. Automatic pollen grain image classification can alleviate the problems of manual categorisation such as subjectivity and time constraints. While a number of ...
Comments