Global Fabric Defect Detection Based on Unsupervised Characterization

doi:10.1007/s12204-020-2246-4

Abstract

Abstract: Fabric texture intelligent analysis comprises the following characteristics: objective detection results, high detection efficiency, and accuracy. It is significantly vital to replace manual inspection for smart green manufacturing in the textile industry, such as quality control and rating, and online testing. For detecting the global image, an unsupervised method is proposed to characterize the woven fabric texture image, which is the combination of principal component analysis (PCA) and dictionary learning. First of all, the PCA approach is used to reduce the dimension of fabric samples, the obtained eigenvector is used as the initial dictionary, and then the dictionary learning method is operated on the defect-free region to get the standard templates. Secondly, the standard templates are optimized by choosing the appropriate dictionary size to construct a fabric texture representation model that can effectively characterize the defect-free texture region, while ineffectively representing the defective sector. That is to say, through the mechanism of identifying normal texture from imperfect texture, a learned dictionary with robustness and discrimination is obtained to adapt the fabric texture. Thirdly, after matching the detected image with the standard templates, the average filter is used to remove the noise and suppress the background texture, while retaining and enhancing the defect region. In the final part, the image segmentation is operated to identify the defect. The experimental results show that the proposed algorithm can adequately inspect fabrics with defects such as holes, oil stains, skipping, other defective types, and non-defective materials, while the detection results are good and the algorithm can be operated flexibly.

Key words: fabric defect detection| unsupervised characterization| fabric texture| learned dictionary

摘要： Fabric texture intelligent analysis comprises the following characteristics: objective detection results, high detection efficiency, and accuracy. It is significantly vital to replace manual inspection for smart green manufacturing in the textile industry, such as quality control and rating, and online testing. For detecting the global image, an unsupervised method is proposed to characterize the woven fabric texture image, which is the combination of principal component analysis (PCA) and dictionary learning. First of all, the PCA approach is used to reduce the dimension of fabric samples, the obtained eigenvector is used as the initial dictionary, and then the dictionary learning method is operated on the defect-free region to get the standard templates. Secondly, the standard templates are optimized by choosing the appropriate dictionary size to construct a fabric texture representation model that can effectively characterize the defect-free texture region, while ineffectively representing the defective sector. That is to say, through the mechanism of identifying normal texture from imperfect texture, a learned dictionary with robustness and discrimination is obtained to adapt the fabric texture. Thirdly, after matching the detected image with the standard templates, the average filter is used to remove the noise and suppress the background texture, while retaining and enhancing the defect region. In the final part, the image segmentation is operated to identify the defect. The experimental results show that the proposed algorithm can adequately inspect fabrics with defects such as holes, oil stains, skipping, other defective types, and non-defective materials, while the detection results are good and the algorithm can be operated flexibly.

关键词: fabric defect detection| unsupervised characterization| fabric texture| learned dictionary

CLC Number:

TS 101.9

WU Ying (吴莹), LOU Lin (娄琳), WANG Jun (汪军). Global Fabric Defect Detection Based on Unsupervised Characterization[J]. J Shanghai Jiaotong Univ Sci, 2021, 26(2): 231-238.

References 22

[1]	JIA L, CHEN C, LIANG J, et al. Fabric defect inspection based on lattice segmentation and Gabor filtering[J]. Neurocomputing, 2017, 238: 84-102.
[2]	SONG A, HAN Y, HU H, et al. A novel texture sensor for fabric texture measurement and classification [J].IEEE Transactions on Instrumentation and Measurement,2014, 63(7): 1739-1747.
[3]	LI B, LI Y H, L¨U Z H. Performance and application of FS220 photoelectric auto cloth inspecting machine[J]. Cotton Textile Technology, 2017, 45(7): 33-36 (in Chinese).
[4]	DONG K, CHEN S, GAO X, et al. Defect detection based on vision system of photoelectric automatic cloth inspection machine [J]. Textile Accessories, 2020,47(1): 57-60 (in Chinese).
[5]	HANBAY K, TALU M F, ¨OZG¨UVEN ¨O F. Fabric defect detection systems and methods: A systematic literature review [J]. Optik, 2016, 127(24): 11960-11973.
[6]	ZHAO C F, CHEN Y, MA J C. Fabric defect detection algorithm based on PHOG and SVM [J]. Indian Journal of Fibre & Textile Research, 2020, 45: 123-126.
[7]	JEYARAJ P R, NADAR E R S. Computer vision for automatic detection and classification of fabric defect employing deep learning algorithm [J]. International Journal of Clothing Science and Technology, 2019,31(4): 510-521.
[8]	DEOTALE N T, SARODE T K. Fabric defect detection adopting combined GLCM, Gabor wavelet features and random decision forest [J]. 3D Research,2019, 10: 5.
[9]	LIU J, WANG C, SU H, et al. Multistage GAN for fabric defect detection [J]. IEEE Transactions on Image Processing, 2020, 29: 3388-3400.
[10]	ZHU Q P, WU M Y, LI J, et al. Fabric defect detection via small scale over-complete basis set [J]. Textile Research Journal, 2014, 84(15): 1634-1649.
[11]	JING J, FAN X, LI P. Patterned fabric defect detection via convolutional matching pursuit dual-dictionary [J].Optical Engineering, 2016, 55(5): 053109.
[12]	WU Y, ZHOU J, AKANKWASA N T, et al. Fabric texture representation using the stable learned discrete cosine transform dictionary [J]. Textile Research Journal,2019, 89(3): 294-310.
[13]	ZHOU J. Automated woven fabric defect detection using dictionary learning [D]. Shanghai, China: Donghua University, 2014 (in Chinese).
[14]	ELAD M, AHARON M. Image denoising via sparse and redundant representations over learned dictionaries[J]. IEEE Transactions on Image Processing, 2006,15(12): 3736-3745.
[15]	GIRYES R, ELAD M. Sparsity-based Poisson denoising with dictionary learning [J]. IEEE Transactions on Image Processing, 2014, 23(12): 5057-5069.
[16]	ZHU N B, TANG T, TANG S, et al. A sparse representation method based on kernel and virtual samples for face recognition [J]. Optik, 2013, 124(23): 6236-6241.
[17]	DONOHO D L, TSAIG Y, DRORI I, et al. Sparse solution of underdetermined systems of linear equations by stagewise orthogonal matching pursuit [J]. IEEE Transactions on Information Theory, 2012, 58(2):1094-1121.
[18]	CHEN S S B, DONOHO D L, SAUNDERS M A.Atomic decomposition by basis pursuit [J]. SIAM Journal on Scientific Computing, 1998, 20(1): 33-61.
[19]	WU Y, WANG J. Woven fabric texture representation and application based on K-SVD dictionary [J].Journal of Textile Research, 2018, 39(2): 165-170 (in Chinese).
[20]	WU Y, WANG J, ZHOU J. Sparse representation of woven fabric texture based on discrete cosine transform over-complete dictionary [J]. Journal of Textile Research, 2018, 39(1): 157-163 (in Chinese).
[21]	LIU G, ZHENG X. Fabric defect detection based on information entropy and frequency domain saliency [J]. Visual Computer, 2020. https://doi.org/10.1007/s00371-020-01820-w (published online).
[22]	ZHOU J, SEMENOVICH D, SOWMYA A, et al. Dictionary learning framework for fabric defect detection[J]. The Journal of the Textile Institute, 2014, 105(3):223-234.