A feature extraction and classification algorithm based on improved sparse auto-encoder for round steel surface defects

Xuguo Yan; Liang Gao; Xuguo Yan; Liang Gao

doi:10.3934/mbe.2020290

Mathematical Biosciences and Engineering

2020, Volume 17, Issue 5: 5369-5394. doi: 10.3934/mbe.2020290

Previous Article Next Article

Research article Special Issues

A feature extraction and classification algorithm based on improved sparse auto-encoder for round steel surface defects

Xuguo Yan ,
Liang Gao ^,

State Key Lab of Digital Manufacturing Equipment and Technology, School of Mechanical Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, China

Received: 28 May 2020 Accepted: 03 August 2020 Published: 12 August 2020

Traditional feature dimensionality reduction (FDR) algorithms can extract features by reducing feature dimensions. However, it may lose some useful information and affect the accuracy of classification. Normally, in traditional defect feature extraction, it first obtain the defect area of the defect image by image preprocessing and defect segmentation, select the original feature set of defects by prior knowledge, and extract the optimal features by traditional FDR algorithms to solve the problem of "curse of dimensionality". In this paper, a feature extraction and classification algorithm based on improved sparse auto-encoder (AE) is proposed. We adopt three traditional FDR algorithms at the same time, combine the defect features obtained in pairs, take the merged defect features as the input of sparse AE, then use the "bottleneck" of sparse AE to conduct the defects classification by Softmax classifier. The experimental results show that the proposed algorithm can extract the optimal features of round steel surface defects with less network training time than individual sparse AE, finally get higher classification accuracy than individual FDR algorithm in the actual production line.
- feature extraction,
- feature dimensionality reduction,
- auto-encoder,
- defect detection,
- round steel
Citation: Xuguo Yan, Liang Gao. A feature extraction and classification algorithm based on improved sparse auto-encoder for round steel surface defects[J]. Mathematical Biosciences and Engineering, 2020, 17(5): 5369-5394. doi: 10.3934/mbe.2020290

Related Papers:

Abstract

Traditional feature dimensionality reduction (FDR) algorithms can extract features by reducing feature dimensions. However, it may lose some useful information and affect the accuracy of classification. Normally, in traditional defect feature extraction, it first obtain the defect area of the defect image by image preprocessing and defect segmentation, select the original feature set of defects by prior knowledge, and extract the optimal features by traditional FDR algorithms to solve the problem of "curse of dimensionality". In this paper, a feature extraction and classification algorithm based on improved sparse auto-encoder (AE) is proposed. We adopt three traditional FDR algorithms at the same time, combine the defect features obtained in pairs, take the merged defect features as the input of sparse AE, then use the "bottleneck" of sparse AE to conduct the defects classification by Softmax classifier. The experimental results show that the proposed algorithm can extract the optimal features of round steel surface defects with less network training time than individual sparse AE, finally get higher classification accuracy than individual FDR algorithm in the actual production line.

References

[1]	B. Wu, T. Xue, T. Zhang, S. H. Ye, A novel method for round steel measurement with a multi-line structure light vision sensor, Meas. Sci. Technol., 21 (2010), 283-293.
[2]	C. I. Brannan, M. A. Bedell, J. L. Resnick, J. J. Eppig, M. A. Handel, D. E. Williams, et al., Developmental abnormalities in Steel17H mice result from a splicing defect in the steel factory sytoplasmic tail, Gene Dev., 10 (1992), 1832-1842.
[3]	Z. Y. Wang, Research on steel plate surface defects detection method based on machine vision, Comput. Modern., 7 (2013), 97-117.
[4]	C. Xie, T. Z. Xie, Key technology of detecting hot heavy rail steel surface faults based on machine vision, J. Chongqing. Univ., 36 (2013), 16-21.
[5]	Q. W. Luo, Y. G. He, A cost-effective and automatic surface defect inspection system for hot-rolled flat steel, Robot Com-int. Manuf., 38 (2016), 16-30.
[6]	F. Meriaudeau, G. Lavallee, E. Fauvet, Machine vision prototype for defect detection on metallic tubes, SPIE- Int. Soc. Opt. Eng., Proc., 4664 (2002), 190-197.
[7]	B. Tang, J. Kong, X. Wang, L. Chen, Steel surface defect recognition based on support vector machine and image processing. China Mechan. Eng., 22 (2011), 1402-1405.
[8]	J. Iiarinen, K. Heikkinen, J. Rauhanmaa, A defect detection scheme for web surface inspection, Int. J. Pattern Recogn., 14 (2000), 735-755.
[9]	J. V. D. Weijer, T. Gevers, A. D. Bagdanov, Boosting color saliency in image feature detection, IEEE T. Pattern Anal., 28 (2005), 150-156.
[10]	Z. W. Qiu, T. W. Zhang, New image color feature extraction method, J. Harbin Ins. Tech., 36 (2004), 1699-1701.
[11]	Y. G. Wang, J. Yang, Y. Zhou, Y. Z. Wang, Region partition and feature matching based color recognition of tongue image, Pattern Recogn. Lett., 28 (2007), 11-19.
[12]	S. S. Patil, A. V. Dusane, Use of color feature extraction technique based on color distribution and relevance feedback for content based image retrieval, Int. J. Comput. Appl., 52 (2012), 9-12.
[13]	W. T. Chen, W. H. Liu, M. S. Chen, Adaptive color feature extraction based on image color distributions, IEEE T. Image Process, 19 (2010), 2005-2016.
[14]	Z. H. Tang, Y. Y. Sun, W. H. Gui, J. P. Liu, Flotation froth image texture feature extraction based on wavelet transform, Comp. Eng., 37 (2011), 206-208.
[15]	C. Y. Pang, J. K. Liu, Improved LFP algorithm on Leukocyte image texture feature extraction and recognition, Acta Photon. Sinica., 42 (2013), 1375-1380.
[16]	Q. Liu, X. P. Liu, L. J. Zhang, L. M. Zhao, Image texture feature extraction & recognition of Chinese herbal medicine based on gray level co-occurrence matrix, Adv. Mater. Res., 605-607 (2012), 2240-2244.
[17]	V. Yu, D. Ruan, D. Nguyen, T. Kaprealian, K. Sheng, SU-F-R-17: Advancing Glioblastoma Multiforme (GBM) recurrence detection with MRI image texture feature extraction and machine learning, Med. Phys., 43 (2016), 3376-3377.
[18]	B. Yuan, B. Xia, D.Zhang, Polarization image texture feature extraction algorithm based on CS-LBP operator, Procedia. Comp. Sci., 131 (2018), 295-301.
[19]	D. X. Wei, X. Y. Chen, R. C. Xu, Image shape feature extraction method based on corner point detection, Comp. Eng., 36 (2010), 220-222.
[20]	O. Mari, N. Seishi, Plant Shape Discrimination of several taxa without shape feature extraction using neural networks with image input, Breed. Sci., 50 (2000), 189-196.
[21]	B. Vijayalakshmi, A new shape feature extraction method for leaf image retrieval, 4th Int. Conf. Signal Image (SIPRO), 221 (2013), 235-245.
[22]	T. Meruliya, P. Dhameliya, J. Patel, D. Panchal, P. Kadam, S. Naik, Image processing for fruit shape and texture feature extraction-review, Int. J. Comput. Appl., 129 (2015), 30-33.
[23]	C. Li, Q. Cao, Extraction method of shape feature for vegetables based on depth image, Trans. Chin. Soc. Agric. Mach., 43 (2012), 242-245.
[24]	L. M. Bruce, C. H. Koger, L. Li, Dimensionality reduction of hyperspectral data using discrete wavelet transform feature extraction, IEEE T. Geosci. Remote, 40 (2002), 2331-2338.
[25]	P. Pudil, J. Novovičová, J. Kittler, Floating search methods in feature selection, Pattern Recogn. Lett., 15 (1994), 1119-1125.
[26]	E. K. Wang, N. Zhe, Y. Li, Z. D. Liang, Y. Ye, A spare deep learning model for privacy attack on remote sensing images, Math. Biosci. Eng., 16 (2019), 1300-1312.
[27]	H. Y. Zhao, C. Che, B. Jin, X. P. Wei, A viral protein identifying framework based on temporal Convolutional network, Math. Biosci. Eng., 16 (2019), 1709-1717.
[28]	Z. C. Zhang, Y. Zhang, T. Zhou, Y. L. Pang, Medical assertion classification in Chinese EMRs using attention enhanced neural network, Math. Biosci. Eng., 16 (2019), 1966-1977.
[29]	E. K. Wang, L. Xi, R. Sun, F. Wang, L. Pan, A new deep learning model for assisted diagnosis on electrocardiogram, Math. Biosci. Eng., 16 (2019), 2481-2491.
[30]	H. J. Deng, L. X. Peng, J. J. Zhang, C. M. Tang, H. L. Fang, H. H. Liu, An intelligent aerator algorithm inspired by deep learning, Math. Biosci. Eng., 16 (2019), 2990-3002.
[31]	S. Y. Chen, Y. Zhang, Y. H. Zhang, J. J. Yu, Y. X. Zhu, Embedded system for road damage detection by deep Convolutional neural network, Math. Biosci. Eng., 16 (2019), 7982-7994.
[32]	Y. X. Zhang, L. C. Jin, B. Wang, D. H. Hu, L. Q. Wang, P. Li, et al., DL-CNV: A deep learning method for identifying copy number variations based on next generation target sequencing, Math. Biosci. Eng., 17 (2020), 202-215.
[33]	Y. Lecun, Y. Bengio, G. Hinton, Deep learning, Nature, 521 (2015), 436.
[34]	L. Z. Wang, S. Q. Guan Strip steel surface defect recognition based on deep learning, J. Xi'an Polytech. Univ., 31 (2017), 669-674.
[35]	L. Yi, G. Li, M. Jiang, An end-to-end steel strip surface defects recognition system based on convolutional neural networks, Steel Res. Int., 87 (2016), 1-5.
[36]	H. H. Hotelling, Analysis of Complex Statistical Variables into Principal Components. Br. J. Educ. Psychol., 24 (1932), 417-520.
[37]	M. T. Kluger, H. Owen, Advantages of PCA exaggerated? Anaesth. Intens. Care, 18 (1990), 588-589.
[38]	M. Rezghi, O. Askar, Noise-free principal component analysis: An efficient dimension reduction technique for high dimensional molecular data, Expert. Syst. Appl., 41 (2014), 7797-7804.
[39]	D. M. Blei, A. Y. Ng, M. I. Jordan, J. Lafferty, Latent dirichlet allocation, J. Mach. Learn. Res., 3 (2003), 993-1022.
[40]	W. Chen, J. E. Meng, S. Q. Wu, PCA and LDA in DCT domain, Pattern Recogn. Lett., 26 (2005), 2474-2482.
[41]	E. Shchepakina, Black swans and canards in self-ignition problem, Nonlin. Anal-real., 4 (2003), 45-50.
[42]	J. N. Wu, J. Wang, L. Liu, Feature extraction via KPCA for classification of gait patterns, Hum. Movem. Sci., 26 (2007), 393-411.
[43]	L. J. Cao, K.S. Chua, W. K. Chong, H. P. Lee, Q. M. Gu, A comparison of PCA, KPCA and ICA for dimensionality reduction in support vector machine, Neurocomputing, 55 (2003), 321-336.
[44]	L. Liu, K. L. Liu, Z. H. Cong, J. L. Zhao, Y. F. Ji, J. He, Long Length Document Classification by Local Convolutional Feature Aggregation, Algorithm., 11 (2018), 109.
[45]	D. Mellado, C. Saavedra, S. Chabert, R. Torres, R. Salas, Self-improving generative artificial neural network for pseudorehearsal incremental class learning, Algorithm, 12 (2019), 206.
[46]	F. Aziz, A. S. W. Wong, S. Chalup, Semi-supervised manifold alignment using parallel deep autoencoders, Algorithms, 12 (2019), 186.

Reader Comments

Your name:*

Email:*
© 2020 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)