An Empirical Study on Fabric Defect Classification Using Deep Network Models
Fabric defect inspection plays an essential role in the textile manufacturing process. Traditional detection is carried out using defect visualization. This method obviously is inefficient in both accuracy and inspection time. Automatic detection, which is based on image processing and machine learning, has been proven to be a suitable approach for this problem. However, due to the variety of defect kinds in a broad range of deferent fabrics, existing methods are actually proposed for typical group of fabric defects. This paper aims to investigate models of deep neural network for the general fabric classification problem. In particular, two models including VGG16 and Darknet are used to classify the defect fabric categories. The models are tested for the TILDA database to evaluate their performances.
KeywordsFabric defect classification Deep neural network Supervised learning
This work is funded by Saigon University, Ho Chi Minh City, Vietnam under the grant number [CS2018-68] (project contract No. 891/HD-QPTKHCN, dated 26/7/2018).
- 4.Hong Kong Productivity Council. Textile Handbook 2000. The Hong Kong Cotton Spinners Association (2000)Google Scholar
- 13.Redmon, J.: Darknet: Open Source Neural Networks in C, 2013–2016. http://pjreddie.com/darknet
- 14.Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale image recognition. CoRR, Vol. abs/1409.1556. http://arxiv.org/abs/1409.1556
- 17.Vikrant, T., Gaurav, S.: Automatic fabric fault detection using morphological operations on bit plane. Int. J. Eng. Res. Technol. 2, 856–861 (2013)Google Scholar
- 22.Zhi, Y.X., Pang, G.K.H., Yung, N.H.C.: Fabric defect detection using adaptive wavelet. In: Proceedings of the IEEE Interntional Conference on Acoustics, Speech, and Signal Processing, ICASSP 2001, pp. 3697–3700, May 2001Google Scholar