Fabric inspection is a crucial process in the textile industry's quality control. Due to the varying structures, textures, geometric features, and spatial distributions of fabric defects, manual fabric inspection is costly and inefficient. Existing fabric defect detection algorithms struggle to strike a balance among efficiency, accuracy, applicability, and deployment costs. In this model, an efficient lightweight fabric defect detection and classification algorithm based on deep convolutional neural networks is proposed.
First, the algorithm performs cluster analysis on the fabric defect dataset to ensure that prior boxes better recall objects with fabric defect geometries and spatial characteristics. Next is fusing the convolutional block attention module attention mechanism and Swin Transformer module with the CSPNet structure. This fusion enhances the model's focus on local features and its ability to capture global contextual information without sacrificing the model's inference speed.
Moreover, WIoU or Wise-IoU is used as the bounding box loss function of the model, which improves the convergence speed of the bounding box loss and enhances the positioning ability of the model. Finally, the performance of the improved model was validated on a public dataset, showing varying degrees of improvement compared to the baseline model and other state-of-the-art algorithms, meeting the requirements of modern textile processes.
MeiSWangYWenG.Automatic fabric defect detection with a multi-scale convolutional denoising autoencoder network model. Sensors2018;
18: 1064.
2.
JingJZhangHWangJ, et al.
Fabric defect detection using Gabor filters and defect classification based on LBP and Tamura method. J Text I2013;
104: 18–27.
3.
LiFYuanLZhangK, et al.
A defect detection method for unpatterned fabric based on multidirectional binary patterns and the gray-level co-occurrence matrix. Text Res J2020;
90: 776–796.
4.
GaoGZhangDLiC, et al. A novel patterned fabric defect detection algorithm based on GHOG and low-rank recovery. In: Baozong Yuan (ed) 2016 IEEE 13th international conference on signal processing (ICSP), Chengdu, China, 06-10 November 2016, pp. 1118–1123. IEEE. New Jersey: Piscataway. DOI: 10.1109/ICSP.2016.7878002
5.
ChenSFengJ. Research on detection of fabric defects based on singular value decomposition. In: Yanling Hao (ed), The 2010 IEEE international conference on information and automation, Chengdu, China, 06-10 November 2016, pp. 857–860.
NJ, Piscataway:
IEEE. DOI: 10.1109/ICINFA.2010.5512449
6.
ChanCPangGK.Fabric defect detection by Fourier analysis. IEEE T Ind Appl2000;
36: 1267–1276.
7.
KimSLeeMHWooK. Wavelet analysis to fabric defects detection in weaving processes. In: IEEE International Symposium on Industrial Electronics, Slovenia Ministrstvo za znanost in tehnologijo (ed) proceedings of the IEEE international symposium on industrial electronics (ISIE'99, Cat. No. 99TH8465), 12-16 July 1999, Bled, Slovenia, pp. 1406–1409.
New York City:
IEEE.
8.
GirshickRDonahueJDarrellT, et al. Rich feature hierarchies for accurate object detection and semantic segmentation. In: IEEE Computer Society Conference on Computer Vision and Pattern Recognition, IEEE Computer Society (ed) proceedings of the IEEE conference on computer vision and pattern recognition, 23-28 June 2014, Columbus, OH, USA, pp. 580–587. Los Alamitos, California: IEEE Computer Society. DOI: 10.1109/CVPR.2014.81
9.
RedmonJDivvalaSGirshickR, et al. You only look once: unified, real-time object detection. In: proceedings of the IEEE conference on computer vision and pattern recognition, Las Vegas, NV, USA, 27-30 June 2016, pp. 779–788. IEEE. DOI: 10.1109/CVPR.2016.91
10.
RonnebergerOFischerPBroxT. U-net: convolutional networks for biomedical image segmentation. In: Nassir Navab, Joachim Hornegger, William M. Wells, Alejandro F. Frangi (eds) medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, 5–9 October 2015, Proceedings, Part III 18 2015, pp. 234–241.
Switzerland:
Springer International Publishing.
11.
ChengLYiJChenA, et al.
Fabric defect detection based on separate convolutional UNet. Multimed Tools Appl2023;
82: 3101–3122.
ChengHLiangJLiuH.Image restoration fabric defect detection based on the dual generative adversarial network patch model. Text Res J2022;
93: 2859–2876.
14.
ZhangHQiaoGLuS, et al.
Attention-based feature fusion generative adversarial network for yarn-dyed fabric defect detection. Text Res J2022;
93: 1178–1195.
15.
ZhangHWangSLuS, et al.
Attention-gate-based u-shaped reconstruction network (AGUR-Net) for color-patterned fabric defect detection. Text Res J2023;
93: 3459–3477.
16.
LuZZhangYXuH, et al.
Fabric defect detection via a spatial cloze strategy. Text Res J2022;
93: 1612–1627.
17.
LiuZTianBLiC, et al.
A context-aware progressive attention aggregation network for fabric defect detection. J. Eng Fiber Fabr2023;
18: 710411732.
18.
WangYXuYYuZ, et al.
Color-patterned fabric defect detection based on the improved YOLOv5s model. Text Res J2023;
93(21-22): 901376195.
19.
ZhouSZhaoJShiYS, et al.
Research on improving YOLOv5s algorithm for fabric defect detection. Int J Cloth Sci Tech2023;
35: 88–106.
20.
OktayOSchlemperJFolgocLL, et al. Attention u-net: learning where to look for the pancreas. arXiv preprint arXiv:1804.039992018.
21.
ZhuXLyuSWangX, et al. TPH-YOLOv5: improved YOLOv5 based on transformer prediction head for object detection on drone-captured scenarios. In: Tamara Berg (ed) proceedings of the IEEE/CVF international conference on computer vision, Montreal, BC, Canada, 11-17 October 2021, pp. 2778–2788.
Piscataway, NJ:
IEEE.
22.
NaSXuminLYongG. Research on k-means clustering algorithm: an improved k-means clustering algorithm. In: Fei Yu (ed) 2010 third international symposium on intelligent information technology and security informatics, Jian, China, 02-04 April 2010, pp. 63–67.
Los Alamitos, Calif:
IEEE Computer Society. DOI: 10.1109/IITSI.2010.74
23.
ArthurDVassilvitskiiS. K-means++ the advantages of careful seeding. In: Proceedings of the Eighteenth Annual ACM-SIAM Symposium on Discrete Algorithms, SODA 2007, New Orleans, Louisiana, USA, 07–09 January, 2007, pp. 1027–1035.
24.
WooSParkJLeeJ, et al. CBAM: convolutional block attention module. In: Vittorio Ferrari, Martial Hebert, Cristian Sminchisescu, Yair Weiss (eds) proceedings of the European conference on computer vision (ECCV), Munich, Germany, 08–14 September, 2018, pp. 3–19. Cham, Switzerland: Springer. DOI: https://doi.org/10.48550/arXiv.1807.06521
LiuZLinYCaoY, et al. Swin transformer: hierarchical vision transformer using shifted windows. In: proceedings of the IEEE/CVF international conference on computer vision, Montreal, QC, Canada, 10-17 October 2021, pp. 10012–10022. Los Alamitos, California: IEEE Computer Society, Conference Publishing Services. DOI: 10.1109/ICCV48922.2021.00986
27.
VaswaniAShazeerNParmarN, et al.
Attention is all you need. Adv Neur Inform Proc Syst2017;
30: 6000–6010.
28.
WangCLiaoHMWuY, et al. CSPNet: a new backbone that can enhance learning capability of CNN. In: proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, Seattle, WA, USA, 14-19 June 2020, pp. 390–391. Los Alamitos, California: IEEE Computer Society, Conference Publishing Services. DOI: 10.1109/CVPRW50498.2020.00203
29.
ZhengZWangPRenD, et al.
Enhancing geometric factors in model learning and inference for object detection and instance segmentation. IEEE T Cybern2021;
52: 8574–8586.
30.
TianDHanYWangS, et al.
Absolute size IoU loss for the bounding box regression of the object detection. Neurocomputing2022;
500: 1029–1040.
31.
TongZChenYXuZ, et al. Wise-IoU: bounding box regression loss with dynamic focusing mechanism. arXiv preprint arXiv:2301.100512023.
WielerMHahnT. Weakly supervised learning for industrial optical inspection. In: DAGM symposium in, 2007.
34.
LiuZHuoZLiC, et al.
DLSE-Net: a robust weakly supervised network for fabric defect detection. Displays2021;
68: 102008.
35.
ZuoLXiaoHWenL, et al.
A pixel-level segmentation convolutional neural network based on global and local feature fusion for surface defect detection. IEEE T Instrum Meas2023;
72: 1–10.
36.
XieHWuZ.A robust fabric defect detection method based on improved RefineDet. Sensors Basel2020;
20: 4260.
37.
XiangZShenYMaM, et al. HookNet: efficient multiscale context aggregation for high-accuracy detection of fabric defects. IEEE T Instrum Meas2023;
72: 1–11.
38.
ZhangYWangWLiZ, et al.
Development of a cross-scale weighted feature fusion network for hot-rolled steel surface defect detection. Eng Appl Artif Intel2023;
117: 105628.
