Fabric Defects Detection Using Multi-Scale Wavelet and Locating

Abstract:

In this paper, multi-scale wavelet edge detection approach is investigated for real time inspection of diversified fabric texture. Multi-scale edge detectors smooth the signal at various scales and detect sharp variations points from their first or second order derivative. The extreme values of the first derivative correspond to the zero crossings of the second derivative and to the inflection points of the smoothed signal. Quadtree decomposition of segmented defects shows pinpoint location of specific web flaw. Further, preliminary graphical user interface (GUI) was designed so as to facilitate operation. After integrate GUI with procedure, parameters material can be acquired, which is vital to applying the inspection system on industrial PC.

• Periodical: Advanced Materials Research (Volume 331)
• Pages: 481-484
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.331.481
• Online since: September 2011
• Authors: Jun Feng Jing, Hang Li, Peng Fei Li
• Keywords: Encoding, GUI, Locating, Multi-Scale Wavelet, Quadtree

References

[1] Ajay Kumar, Grantham K.H. Pang: IEEE Transactions on Industry Applications, (2002),p.425

[2] Li-Wei Han, De Xu: International Journal of Automation and Computing, (2010),p.86

[3] Jiri Kula, Maros Tunak, Ales Linka: 7th International Conference, Liberec, Czech Republic (2010).

[4] Zhang Yu: Automatic Fabric Defect Detection and Classification Using Machine Learning Technology (Xi'an,China,2010).

[5] P. Hornby, F. Boschetti and F.G. Horowitz.: Analysis of Potential Field Data in The Wavelet Domain (Australia)