Texture analysis and defect classification for fabric images using regular bands and quadratic programming

Abstract

Defect detection is a key problem in quality control for many industrial fields like wallpaper scanning, ceramic flow detection and fabric inspection. For a long time the fabric defects inspection process is still carried out with human visual inspection, and thus, insufficient and costly. Therefore, automatic fabric defect inspection is required to reduce the cost and time waste caused by defects. Many techniques have been developed for detection of defects for fabrics through the years using neural networks, Fourier transform. However, most of the methods mentioned above are mainly designed for un-patterned fabric inspection. In this paper, the work is concentrated on the patterned texture inspection of the fabrics, using regular bands and enhancement of these images using linear quadratic programming. © 2013 Springer.