In the textile industry, detecting defects in fabrics featuring intricate patterns and small imperfections poses a persistent challenge. There is an urgent need for defect detection systems characterized by high accuracy and ease of deployment. To address these needs, a fabric defect detection algorithm, SR-NET, has been proposed in this paper, featuring lightweight and enhanced multiple feature extraction capabilities. The Spatial Receptive-field Convolution (SRConv) module was introduced as a convolutional module, effectively optimizing the performance of the convolutional neural network and improving overall model detection performance. The Region-Semantic Residual Module (RSR) module was proposed as an attention mechanism, which was fused with the C2f module to form the C2f_RSR module. This enhancement significantly enhanced the modeling ability to extract features pertinent to small defects. To reduce the sensitivity of the model to deviations in the position of small objects, The Gaussian Wasserstein Distance (GWD) metric was introduced into LGWD as a loss function. This adjustment has demonstrated outstanding performance in small target detection. Experimental results demonstrated that SR-NET enhanced mAP by 5.6% compared with baseline model. Simultaneously, SR-NET kept parameters and FLOPs almost unchanged. SR-NET proved to be well suited for practical production applications, and is able to meet real-time detection requirements.
CevheriCI.The effect of rain on fiber and yarn quality of cotton (G. hirsutum L.) varieties in two different growing seasons. J Nat Fibers2022;
19(9): 3558–3570.
2.
Association of the Hong Kong Cotton Spinners. Textile hand book.
Hong Kong:
The Hong Kong Cotton Spinners Association, 2001, pp. 841–844.
3.
SrinivasanKDastoorPHRadhakrishnaiahP, et al.
FDAS: A knowledge-based framework for analysis of defects in woven textile structures. J Textile Inst1992;
83(3): 431–448.
WangQCJingJFZhangL, et al.
Denim defect detection based on optimal Gabor filter. Laser Optoelectron Prog2018;
55(7): 363–369.
6.
DongSWangPAbbasK.A survey on deep learning and its applications. Comput Sci Rev2021;
40: 100379.
7.
RenZFangFYanN, et al.
State of the art in defect detection based on machine vision. Int J Precis Eng Manuf Green Technol2022;
9(2): 661–691.
8.
RasheedAZafarBRasheedA, et al.
Fabric defect detection using computer vision techniques: A comprehensive review. Math Probl Eng2020;
2020: 1–24.
9.
KuuttiSBowdenRJinY, et al.
A survey of deep learning applications to autonomous vehicle control. IEEE Trans Intell Transp Syst2020;
22(2): 712–733.
10.
RedmonJDivvalaSGirshickR, et al. You only look once: Unified, real-time object detection. Proceedings of the IEEE conference on computer vision and pattern recognition. 2016, pp. 779–788.
11.
RedmonJFarhadiA. YOLO9000: better, faster, stronger. Proceedings of the IEEE conference on computer vision and pattern recognition. 2017, pp. 7263–7271.
12.
RedmonJFarhadiA. Yolov3: An incremental improvement. arXiv preprint, arXiv:1804.02767, 2018.
13.
AdarshPRathiPKumarM. “YOLO v3-Tiny: Object Detection and Recognition using one stage improved model,” 2020 6th International Conference on Advanced Computing and Communication Systems (ICACCS), Coimbatore, India, 2020, pp. 687–694. doi: 10.1109/ICACCS48705.2020.9074315.
14.
BochkovskiyAWangCYLiaoHYM. Yolov4: Optimal speed and accuracy of object detection. arXiv preprint, arXiv:2004.10934, 2020.
LiCLiLJiangH, et al. YOLOv6: A single-stage object detection framework for industrial applications. arXiv preprint, arXiv:2209.02976, 2022.
17.
WangCYBochkovskiyALiaoHYM. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 7464–7475.
ZhangXLiuCYangD, et al. RFAConv: Innovating Spatital Attention and Standard Convolutional Operation. arXiv preprint, arXiv:2304.03198, 2023.
22.
WeiHLiuXXuS, et al. DWRSeg: Dilation-wise Residual Network for Real-time Semantic Segmentation. arXiv preprint, arXiv:2212.01173, 2022.
23.
WangJXuCYangW, et al. A normalized Gaussian Wasserstein distance for tiny object detection. arXiv preprint, arXiv:2110.13389, 2021.
24.
DhivyaMRenuka-DeviM.Detection of structural defects in fabric parts using a novel edge detection method. Comput J2018;
62(7): 1036–1043.
25.
WuYZhouJAkankwasaNT, et al.
Fabric texture representation using the stable learned discrete cosine transform dictionary. Text Res J2019;
89(3): 294–310.
26.
DongRLiB.Fabric defect detection algorithm based on optimal parametric nonlinear GLCM. Comput Eng Des2015;
36(09): 2467–2471+2507. DOI: 10. 16208/j. issn1000-7024.2015.09.029
27.
LiuXXueFTengL. Surface defect detection based on gradient LBP. 2018 IEEE 3rd International Conference on Image, Vision and Computing (ICIVC), Chongqing, China, 2018, pp. 133–137, doi: 10.1109/ICIVC.2018.8492798.
28.
LiMWanSDengZ, et al.
Fabric defect detection based on saliency histogram features. Comput Intell2019.
29.
WangZYZhaoXJYangM.Towel fabric defect detection based on frequency domain analysisComputMeasure Control2020;
28(06): 28–31t55. DOI:10.16526/j.cnki.11-4762/tp.2020.06.006
30.
ShiK. Objective evaluation of fabric flatness grade based on image frequency domain analysis. Jiangnan University, 2020. DOI:10.27169/d.cnki.gwqgu.2020.000985
31.
ZhaoSXYinLZangJ, et al.
Real-time fabric defect detection based on multi-scale convolutional neural network [J]. IET Collab Intell Manuf2020;
2(4): 189–196.
32.
MengAZhengLMWangSY, et al.
An improved Faster R-CNN method for fabric defect detection. Small Microcomput Syst2021;
42(5): 1029–1033.
33.
JinRNiuQ.Automatic fabric defect detection based on an improved YOLOv5[J/OL]. Math Probl Eng2021; 1–13.
34.
YangJJinYXLiuYB, et al.
Research on fabric material recognition method based on improved Transformer. J North Central Uni (Nat Sci Ed)2023;
44(2): 138–145, 161.
MabeJZubiaJGorritxategiE.Photonic low cost micro-sensor for in-line wear particle detection in flowing lube oils. Sensors2017;
17(3): 586.
37.
AberCHamidAElchikhM, et al.
Eddy current microsensor and RBF neural networks for detection and characterization of small surface defects. Measure Sci Rev2022;
22(3): 112–121.
LiuSHuangD. Receptive field block net for accurate and fast object detection. Proceedings of the European conference on computer vision (ECCV), 2018, pp. 385–400.
40.
ChuXLiLZhangB. Make RepVGG greater again: A quantization-aware approach. arXiv preprint, arXiv:2212.01593, 2022.
41.
XuXJiangYChenW, et al. Damo-yolo: A report on real-time object detection design. arXiv preprint, arXiv:2211.15444, 2022.
42.
ZhangQLYangYB. Sa-net: Shuffle attention for deep convolutional neural networks. ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2021, pp. 2235–2239.
43.
MisraDNalamadaTArasanipalaiAU, et al. Rotate to attend: Convolutional triplet attention module. Proceedings of the IEEE/CVF winter conference on applications of computer vision. 2021, pp. 3139–3148.
44.
LiXWangWHuX, et al. Selective kernel networks. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 510–519.
45.
HuJShenLSunG. Squeeze-and-excitation networks. Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 7132–7141.
46.
LiuYShaoZHoffmannN. Global attention mechanism: Retain information to enhance channel-spatial interactions. arXiv preprint, arXiv:2112.05561, 2021.
47.
VaswaniAShazeerNParmarN, et al.
Attention is all you need. Adv Neural Inf Process Syst2017, 30.
48.
YuZHuangHChenW, et al. Yolo-facev2: A scale and occlusion aware face detector. arXiv preprint, arXiv:2208.02019, 2022.
49.
ZhangHXuCZhangS. Inner-IoU: More Effective Intersection over Union Loss with Auxiliary Bounding Box. arXiv preprint, arXiv:2311.02877, 2023.
50.
ZhangYFRenWZhangZ, et al.
Focal and efficient IOU loss for accurate bounding box regression. Neurocomputing2022;
506: 146–157.
51.
TongZChenYXuZ, et al. Wise-IoU: Bounding box regression loss with dynamic focusing mechanism. arXiv preprint, arXiv:2301.10051, 2023.
52.
RenSHeKGirshickR, et al.
Faster r-cnn: Towards real-time object detection with region proposal networks. Adv Neural Inf Process Syst2015, 28.
53.
HeKGkioxariGDollárP, et al. Mask r-cnn. Proceedings of the IEEE international conference on computer vision, 2017, pp. 2961–2969.
54.
Zhang H, Li F, Liu S, et al. Dino: Detr with improved denoising anchor boxes for end-to-end object detection[J]. arXiv preprint arXiv:2203.03605, 2022.
55.
Lyu C, Zhang W, Huang H, et al. Rtmdet: An empirical study of designing real-time object detectors[J]. arXiv preprint arXiv:2212.07784, 2022.
56.
HanKWangYTianQ, et al. Ghostnet: More features from cheap operations. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 1580–1589.
57.
HowardASandlerMChuG, et al. Searching for mobilenetv3. Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp.1314–1324.
58.
LiXWangWWuL, et al.
Generalized focal loss: Learning qualified and distributed bounding boxes for dense object detection. Adv Neural Inf Process Syst2020;
33: 21002–21012.
