Visual Property of Textile Design

Part 4 : Extraction of Textile Defects Using Wavelet Transform

Abstract

In textile, the patterns of the surface may be due to either the yarn's shapes or the repetition of the warp and weftyarn. The regular periodic nature of many textile patterns permits Wavelet transform techniques in image processing tobe used to measure their visual characteristics. The Fourier power spectrum provides a useful description of the spatialfrequency content in a digital image, and in particular the coarseness of any texture present, but loses the local information.The Wavelet transform can obtain both the frequency information and the local information. The Wavelet transformfixed by frequency of fabric density can graphically describe the irregular part of frequency components on textile surfaceand gives a means of extracting the textile defects.