The Automating Detecting of Stitch Distortion in Knitted Fabric by Image Processing Technology

Abstract

The stitch distortion of knitted fabric is an important parameter for the fabric detecting and it is measured manually at present. The paper discusses how to process knitted fabric image and obtain the stitch distortion angle automatically using image processing technology. By rotational property of Fourier transform, i.e. if a function f(x, e) of the image turns an angle e in the space domain, the Fourier transform of the function would turn a same angle in the transform frequency domain, that is, if the angle between stitch row and course in a knitted fabric image is e, the angle between the lines of their feature points in their transform frequency domain is also the same, an automatic detecting system was established and the angle of stitch distortion was been obtained automatically. Through examination verifying, the detecting result is credible and the automatic detecting of the stitch distortion is realized.