The lack of retrieval methods in the textile subdivision field makes it difficult for universal retrieval methods to meet the precision requirements of plaid fabric manufacturing enterprises. Based on the characteristics of plaid fabrics, this paper presents a novel image retrieval method for plaid fabrics using a convolutional neural network (CNN) with an attention mechanism. Global and local deep features are extracted by combining the attention mechanism with the CNN branches to fully characterize the plaid fabric images, which are then fused by an orthogonal fusion module. To reduce the amount of training data, a novel training strategy is designed to optimize the feature extraction and fusion tasks. The Annoy algorithm is used as the similarity measurement method to balance the retrieval precision and efficiency. To verify the proposed scheme, over 44,000 fabric samples are collected from the factory to build the image database as the benchmark. Experiments show that precision and recall at rank five are up to 77.5% and 57.1%, respectively, and the mean average precision is up to 0.758. Results prove that the proposed method is effective and efficient; it can provide references for fabric manufacturing and exploit the advantages of historical production experience.
LiWDuanLXuD, et al. Text-based image retrieval using progressive multi-instance learning. In: 2011 international conference on computer vision, Barcelona, Spain, 6–13 November 2011, pp. 2049–2055. Piscataway, NJ: IEEE. DOI: 10.1109/ICCV.2011.6126478.
2.
VishrajRGuptaS, andSinghS.A comprehensive review of content-based image retrieval systems using deep learning and hand-crafted features in medical imaging: Research challenges and future directions. Comput Electr Eng2022;
104: 108450. DOI: 10.1016/j.compeleceng.2022.108450.
3.
ZhangNXiangJWangL, et al.
Research progress of content-based fabric image retrieval. Text Res J2023;
93: 1401–1418. DOI: 10.1177/00405175221128524.
4.
ZhangLLiuXLuZ, et al. Lace fabric image retrieval based on multi-scale and rotation invariant LBP. In: Proceedings of the 7th international conference on internet multimedia computing and service (ed Jain R, Jiang S, Smith J et al.), Zhangjiajie, China, 19–21 August 2015, paper no. 74. New York: Association for Computing Machinery. DOI: 10.1145/2808492.2808567.
5.
ZhangNXiangJWangL, et al.
Image retrieval of wool fabric. Part I: Based on low-level texture features. Text Res J2019;
89: 4195–4207. DOI: 10.1177/0040517519829003.
6.
ZhangNXiangJWangL, et al.
Image retrieval of wool fabric. Part II: Based on low-level color features. Text Res J2020;
90: 797–808. DOI: 10.1177/0040517519881819.
7.
LiYLuoHJiangG, et al.
Content-based lace fabric image retrieval system using texture and shape features. J Text Inst2019;
110: 911–915. DOI: 10.1080/00405000.2018.1532782.
8.
JingJLiQLiP, et al.
A new method of printed fabric image retrieval based on color moments and gist feature description. Text Res J2016;
86: 1137–1150. DOI: 10.1177/0040517515606378.
9.
SuciatiNHerumurtiD, andWijayaAY. Fractal-based texture and HSV color features for fabric image retrieval. In: 2015 IEEE international conference on control system, computing and engineering (ICCSCE), Penang, Malaysia, 27–29 November 2015, pp. 178–182. Piscataway, NJ: IEEE. DOI: 10.1109/ICCSCE.2015.7482180.
10.
DengDWangRWuH, et al.
Learning deep similarity models with focus ranking for fabric image retrieval. Image Vision Comput2018;
70: 11–20. DOI: 10.1016/j.imavis.2017.12.005.
11.
ZhangNShameyRXiangJ, et al.
A novel image retrieval strategy based on transfer learning and hand-crafted features for wool fabric. Expert Syst Appl2022;
191: 116229. DOI: 10.1016/j.eswa.2021.116229.
12.
WangMWangJZhangN, et al.
Fabric image retrieval based on decoupling of texture and color feature. J Eng Fibers Fabr. Epub ahead of print 20 May 2024. DOI: 10.1177/15589250241246074.
13.
YangMHeDFanM, et al. DOLG: Single-stage image retrieval with deep orthogonal fusion of local and global features. In: Proceedings of the 2021 IEEE/CVF international conference on computer vision (ICCV), Montreal, QC, Canada, 10–17 October 2021, pp. 11752–11761. Piscataway, NJ: IEEE. DOI: 10.1109/ICCV48922.2021.01156.
14.
LiXMaJ.Fine-grained image retrieval by combining attention mechanism and context information. Neural Comput Appl2023;
35: 1881–1897. DOI: 10.1007/s00521-022-07873-3.
15.
ZhangHXiongWLuS, et al.
QA-USTNet: Yarn-dyed fabric defect detection via U-shaped Swin Transformer Network based on Quadtree Attention. Text Res J2023;
93: 3492–3508. DOI: 10.1177/00405175231158134.
16.
ZhangHLiuSLuS, et al.
Knowledge distillation for unsupervised defect detection of yarn-dyed fabric using the system DAERD: Dual attention embedded reconstruction distillation. Color Technol2024;
140: 125–143. DOI: 10.1111/cote.12705.
17.
ZhangHLiuSWangC, et al.
Color-patterned fabric defect detection algorithm based on triplet attention multi-scale U-shape denoising convolutional auto-encoder. J Supercomput2024;
80: 4451–4476. DOI: 10.1007/s11227-023-05639-9.
18.
KibriyaAMFrankE. An empirical comparison of exact nearest neighbour algorithms. In: Knowledge discovery in databases: PKDD 2007 (eds Kok JN, Koronacki J, Lopez de Mantaras R, et al.), Berlin, Heidelberg. Berlin: Springer, 2007, pp. 140–151.
19.
AumüllerMBernhardssonE, andFaithfullA.ANN-Benchmarks: A benchmarking tool for approximate nearest neighbor algorithms. Inf Syst2020;
87: 101374. DOI: 10.1016/j.is.2019.02.006.
20.
BentleyJL.Multidimensional binary search trees used for associative searching. Commun ACM1975;
18: 509–517. DOI: 10.1145/361002.361007.
MujaMLoweDG. Fast approximate nearest neighbors with automatic algorithm configuration. In: International conference on computer vision theory and applications, Lisbon, Portugal, 5–8 February 2009, vol. 1, pp. 331–340. Lisbon: INSTICC Press.
23.
MalkovYAYashuninDA.Efficient and robust approximate nearest neighbor search using hierarchical navigable small world graphs. IEEE Trans Pattern Anal Mach Intell2020;
42: 824–836. DOI: 10.1109/TPAMI.2018.2889473.
24.
ZhangXGaoW, andPanR.Deep neural network with strip pooling for image classification of yarn-dyed plaid fabrics. Comp Model Eng Sci2022;
130: 1533. DOI: 10.32604/cmes.2022.018763.
25.
ZhangNXiangJWangL, et al.
Image retrieval of wool fabric. Part III: Based on aggregated convolutional descriptors and approximate nearest neighbors search. Text Res J2022;
92: 434–445. DOI: 10.1177/00405175211037186.
26.
DengJGuoJYangJ, et al.
ArcFace: Additive angular margin loss for deep face recognition. IEEE Trans Pattern Anal Mach Intell2022;
44: 5962–5979. DOI: 10.1109/TPAMI.2021.3087709.
27.
HameedIMAbdulhussainSH, andMahmmodBM.Content-based image retrieval: A review of recent trends. Cogent Eng2021;
8: 1927469. DOI: 10.1080/23311916.2021.1927469.
28.
SmeuldersAWMWorringMSantiniS, et al.
Content-based image retrieval at the end of the early years. IEEE Trans Pattern Anal Mach Intell2000;
22: 1349–1380. DOI: 10.1109/34.895972.
29.
BhatJIYousufRJeelaniZ, et al. An insight into content-based image retrieval techniques, datasets, and evaluation metrics. In: SheikhJAKhanTKanaujiaBK (eds) Intelligent Signal Processing and RF Energy Harvesting for State of art 5G and B5G Networks.
Singapore:
Springer Nature, 2024, pp. 127–146.
30.
DubeySRSinghSK, andChuWT. Vision transformer hashing for image retrieval. In: 2022 IEEE international conference on multimedia and expo (ICME), Taipei, Taiwan, 18–22 July 2022. Piscataway, NJ: IEEE.
