Fabric defects detecting and rank scoring based on Fisher criterion discrimination

• Published in: 2009 International Conference on Machine Learning and Cybernetics Vol. 5; pp. 2560 - 2563
• Main Authors: Sheng-Wang Li, Li-Wei Guo, Chun-Hua Li
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2009
• Subjects: Application software; Auto correlation function; Autocorrelation; Computer applications; Fabric defect detection; Fabric defect rating; Fabrics; Fisher criterion discrimination; Image processing; Image segmentation; Inspection; Pattern recognition; Pixel; Surface texture; Textiles
• ISBN: 9781424437023; 1424437024
• ISSN: 2160-133X
• DOI: 10.1109/ICMLC.2009.5212106

Automatic texture defect detection is highly important for many fields of visual inspection. This paper studies the application of advanced computer image processing techniques for solving the problem of automated defect detection for textile fabrics. The approach is used for the quality inspection of local defects embedded in homogeneous textured surfaces. Above all, the size of the basic texture units of the fabric image is acquired by calculating auto correlation function in weft direction and in wrap direction. Then the sizes of the basic texture units are taken as criterion to segment the fabric image. During scanning the fabric texture image, the basic units are segmented. And the Fisher criterion discriminator is used to assign each unit to a class at the same time. Afterwards, the fabric detects are measured according to the relationship of the suffix of the image pixel and the scale of the image and ranked scale by comparing with America Four Points System. Experiments with real fabric image data show that it is effective.