Abstract
Anomaly detection in textile images poses significant challenges due to the scarcity of defective samples and the complex nature of textile textures. This study presents a novel image processing workflow that enhances the unsupervised Variational Autoencoder’s (VAE) ability to identify anomalies in textile images, addressing the limitation of insufficient defective samples in real-world manufacturing scenarios. The primary motivation behind this research is to develop a robust anomaly detection method that can be trained using only normal samples, overcoming the common imbalance between normal and defective samples in the textile industry. Our proposed method introduces domain-specific techniques to preprocess images, assess the adequacy of training samples, and employ intuitive visual methods to differentiate between normal and abnormal samples. A key strength of our approach lies in strategically crop** original images into smaller blocks, increasing training samples and computational efficiency. However, this crop** step introduces abrupt boundary issues that can hinder accurate anomaly detection. To mitigate this problem, we developed a refined image processing approach that effectively resolves boundary artifacts, enabling precise localization of abnormal regions. We trained, tested, and validated our VAE model using the TILDA textile texture database. The experimental results highlight the robustness of our method, achieving high identification rates of 74% for normal samples and 76.9% for abnormal samples, even when trained solely on normal samples. The insights gained from this study have significant implications for the textile industry, paving the way for more efficient and reliable quality control processes.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Data availability
No datasets were generated or analysed during the current study.
Abbreviations
- CNN:
-
Convolutional Neural Network
- GPU:
-
Graphics Processing Units
- Faster R-CNN:
-
Faster Region with Convolution Neural Network
- YOLOV2:
-
You Only Look Once V2
- YOLOV3:
-
You Only Look Once V3
- YOLOV4:
-
You Only Look Once V4
- GAN:
-
Generative Adversarial Networks
- FCSDA:
-
Fisher Criterion-based Stacked Denoising Autoencoders
- MS-FCAE:
-
Multiscale Feature-clustering-based Fully Convolutional Autoencoder
- DAE:
-
Denoising AutoEncoder
- SAE:
-
Sparse AutoEncoder
- CVAE:
-
Contractive AutoEncoder
- VAE:
-
Variational Autoencoder
- TILDA:
-
Textile Texture-Database
- VAE:
-
Variational Autoencoder
- AE:
-
Autoencoder
- ELBO:
-
Evidence Lower Bound
- t-SNE:
-
t-Distributed Stochastic Neighbor Embedding
References
Asperti A, Trentin M (2020) Balancing Reconstruction Error and Kullback-Leibler divergence in Variational Autoencoders. Ieee Access 8:199440–199448. https://doi.org/10.1109/ACCESS.2020.3034828
Belkina AC, Ciccolella CO, Anno R, Halpert R, Spidlen J, Snyder-Cappione JE (2019) Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets. Nat Commun 10(1):5415. https://doi.org/10.1038/s41467-019-13055-y
Bui AT, Apley DW (2018) Monitoring for changes in the nature of stochastic textured surfaces. J Qual Technol 50(4):363–378. https://doi.org/10.1080/00224065.2018.1507559
Chen X, Yu J, Kong S, Wu Z, Fang X, Wen L (2019) Towards Real-Time Advancement of Underwater Visual Quality with GAN. IEEE Trans Industr Electron 66(12):9350–9359. https://doi.org/10.1109/TIE.2019.2893840
Chen TT, Liu XP, **a BZ, Wang W, Lai YZ (2020) Unsupervised anomaly detection of Industrial Robots using sliding-window Convolutional Variational Autoencoder. IEEE Access 8:47072–47081. https://doi.org/10.1109/access.2020.2977892
Guarnera GC, Hall P, Chesnais A, Glencross M (2017) Woven Fabric Model Creation from a single image. Acm Trans Graphics 36(5) Article 165. https://doi.org/10.1145/3132187
Jian BL, Chen CL, Chu WL, Huang MW (2017) The facial expression of schizophrenic patients applied with infrared thermal facial image sequence. BMC Psychiatry 17:229. https://doi.org/10.1186/s12888-017-1387-y
Karamti H, Lashin MMA, Alrowais FM, Mahmoud AM (2021) A New Deep Stacked Architecture for Multi-fault Machinery Identification with Imbalanced samples. Ieee Access 9:58838–58851. https://doi.org/10.1109/Access.2021.3071796
Kingma DP, Welling M (2013) Auto-Encoding Variational Bayes. ar**v:1312.6114. Retrieved December 01, 2013, from https://ui.adsabs.harvard.edu/abs/2013ar**v1312.6114K
Kuo CFJ, Jian BL, Wu HC, Peng KC (2012) Automatic machine embroidery image color analysis system. Part I: using Gustafson-Kessel clustering algorithm in embroidery fabric color separation. Text Res J 82(6):571–583. https://doi.org/10.1177/0040517511435012
Lafarge MW, Caicedo JC, Carpenter AE, Pluim JPW, Singh S, Veta M (2019) Capturing single-cell phenotypic variation via unsupervised representation learning. Proc Mach Learn Res 103:315–325
Lee S, Kwak M, Tsui KL, Kim SB (2019) Process monitoring using variational autoencoder for high-dimensional nonlinear processes. Eng Appl Artif Intell 83:13–27. https://doi.org/10.1016/j.engappai.2019.04.013
Li JH, Struzik Z, Zhang LQ, Cichocki A (2015) Feature learning from incomplete EEG with denoising autoencoder [Article]. NEUROCOMPUTING, 165, 23–31. https://doi.org/10.1016/j.neucom.2014.08.092
Li ZH, Fan YY, Liu WH, Yu ZQ, Wang FQ (2017b) Emotional textile image classification based on cross-domain convolutional sparse autoencoders with feature selection. J Electron Imaging 26(1). https://doi.org/10.1117/1.Jei.26.1.013022
Martin R, Xue M, Klein R, Hullin MB, Weinmann M (2019) Using patch-based image synthesis to measure perceptual texture similarity. COMPUTERS GRAPHICS-UK 81:104–116. https://doi.org/10.1016/j.cag.2019.04.001
Mei S, Wang YD, Wen GJ (2018) Automatic fabric defect detection with a Multi-scale Convolutional Denoising Autoencoder Network Model. Sensors 18(4):Article1064. https://doi.org/10.3390/s18041064
Nalepa J, Marcinkiewicz M, Kawulok M (2019) Data augmentation for brain-tumor segmentation: a review. Front Comput Neurosci 13:83. https://doi.org/10.3389/fncom.2019.00083
Nsengiyumva P (2014) A vision-based Quality Inspection System for Fabric Defect Detection and classification. Central University of Technology, Free State], Bloemfontein
Odaibo S (2019) J. a. e.-p. Tutorial: Deriving the Standard Variational Autoencoder (VAE) Loss Function. ar**v:1907.08956. Retrieved July 01, 2019, from https://ui.adsabs.harvard.edu/abs/2019ar**v190708956O
Pawar K, Attar VZ (2020) Assessment of Autoencoder Architectures for Data Representation. In W. Pedrycz & S.-M. Chen (Eds.), Deep Learning: Concepts and Architectures (pp. 101–132). Springer International Publishing. https://doi.org/10.1007/978-3-030-31756-0_4
Redmon J, Farhadi A (2016) YOLO9000: Better, Faster, Stronger. ar**v:1612.08242. Retrieved December 01, 2016, from https://ui.adsabs.harvard.edu/abs/2016ar**v161208242R
Redmon J, Farhadi A (2018) YOLOv3: An Incremental Improvement. ar**v:1804.02767. Retrieved April 01, 2018, from https://ui.adsabs.harvard.edu/abs/2018ar**v180402767R
Ren SQ, He KM, Girshick R, Sun J (2017) Faster R-CNN: towards real-time object detection with region proposal networks. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE. 39(6):1137–1149. https://doi.org/10.1109/tpami.2016.2577031
Sekulska-Nalewajko J, Goclawski J, Korzeniewska E (2020) A method for the Assessment of Textile Pilling Tendency using Optical Coherence Tomography. Sensors 20(13):Article3687. https://doi.org/10.3390/s20133687
Sengupta S, Basak S, Saikia P, Paul S, Tsalavoutis V, Atiah F, Ravi V, Peters A (2020) A review of deep learning with special emphasis on architectures, applications and recent trends. Knowl Based Syst 194:105596. https://doi.org/10.1016/j.knosys.2020.105596
Tabernik D, Sela S, Skvarc J, Skocaj D (2020) Segmentation-based deep-learning approach for surface-defect detection [Article]. J Intell Manuf 31(3):759–776. https://doi.org/10.1007/s10845-019-01476-x
Umirzakova S, Ahmad S, Mardieva S, Muksimova S, Whangbo TK (2023) Deep learning-driven diagnosis: a multi-task approach for segmenting stroke and Bell’s palsy. Pattern Recogn 144:109866. https://doi.org/10.1016/j.patcog.2023.109866
Van Eycke YR, Foucart A, Decaestecker C (2019) Strategies to reduce the Expert Supervision required for deep learning-based segmentation of histopathological images. Front Med 6:Article222. https://doi.org/10.3389/fmed.2019.00222
Wang M, Chapman L, Moore M, Suh M (2019) Replication of screen-Printing Fabric via ink-jet Textile Printing. J Imaging Sci Technol 63(4) Article 040402. https://doi.org/10.2352/J.ImagingSci.Technol.2019.63.4.040402
Wei XK, Yang ZM, Liu YX, Wei DH, Jia LM, Li YJ (2019) Railway track fastener defect detection based on image processing and deep learning techniques: a comparative study [Article]. Eng Appl Artif Intell 80:66–81. https://doi.org/10.1016/j.engappai.2019.01.008
Wei B, Hao KR, Gao L, Tang XS (2020) Detecting textile micro-defects: a novel and efficient method based on visual gain mechanism. Inf Sci 541:60–74. https://doi.org/10.1016/j.ins.2020.06.035
Yang H, Chen YF, Song KY, Yin ZP (2019) Multiscale feature-clustering-based fully convolutional autoencoder for fast Accurate Visual Inspection of texture surface defects [Article]. IEEE Trans Autom Sci Eng 16(3):1450–1467. https://doi.org/10.1109/Tase.2018.2886031
Zhu JL, Jiang MY, Liu Z (2022) Fault Detection and Diagnosis in industrial processes with Variational Autoencoder: a Comprehensive Study. Sensors 22(1):Article227. https://doi.org/10.3390/s22010227
Acknowledgements
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded in part by the National Science and Technology Council, Taiwan, ROC under Grants NSTC 109-2221-E-167 -010 and NSTC 110-2221-E-167 -017.
Author information
Authors and Affiliations
Contributions
Bo-Lin Jian was the primary manuscript writer and was responsible for designing the initial structure of the article. Wen-Lin Chu contributed to revising the article structure and provided the conceptual ideas for the Variational Autoencoder (VAE) implementation. Qun-Wei Chang assisted in organizing the manuscript and supported the experimental data collection process.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Conflicting interests
The Author declare that there is no conflict of interest.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Chu, WL., Chang, QW. & Jian, BL. Unsupervised anomaly detection in the textile texture database. Microsyst Technol (2024). https://doi.org/10.1007/s00542-024-05711-1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00542-024-05711-1