Defect Detection in Textiles with Co-occurrence Matrix as a Texture Model Description
Automatized inspection at textile production lines becomes very important. However, there is still a need to design methods which meet not only demands concerning accuracy of defect detection, but also ones related to the processing time. In this work, a novel approach for defect model definition is presented. It is derived from the idea of co-occurrence matrix. Due to scale incorporation and binarization of the model content it proved to be a very powerful descriptor of the novelties. Moreover, it also satisfies the requirements of short processing time. The defect mask achieved with the introduced method was compared visually to other popular solutions and show a very high accuracy and quality of defect description. The processing time is real-time as the response for a 1MP (megapixel) image is reached within tens of milliseconds.
KeywordsDefect detection Image segmentation Co-occurrence matrix
This work has been based on the results of the project “Opracowanie systemu do efektywnej integracji aplikacji wizyjnych przez użyt- kowników końcowych” co-financed by the European Regional Development Fund under Operational Programme Innovative Economy 2007–2013, based on the Agreement no. UDA-POIG.01.04.00-24-067/11-00.
- 10.Iyer, M., Janakiraman, S.: Defect detection in pattern texture analysis. In: 2014 International Conference on Communication and Signal Processing, pp. 172–175, April 2014Google Scholar
- 13.Nurzynska, K., Kubo, M., Muramoto, K.: Snow particle automatic classification with texture operators. In: 2011 IEEE International Geoscience and Remote Sensing Symposium, pp. 2892–2895, July 2011Google Scholar
- 18.Sari, L., Ertüzün, A.: Texture defect detection using independent vector analysis in wavelet domain. In: 2014 22nd International Conference on Pattern Recognition, pp. 1639–1644, August 2014Google Scholar
- 19.Schölkopf, B., Williamson, R., Smola, A., Shawe-Taylor, J., Platt, J.: Support vector method for novelty detection. In: Proceedings of the 12th International Conference on Neural Information Processing Systems, NIPS 1999, pp. 582–588. MIT Press, Cambridge (1999)Google Scholar