IMAGE-ANALYSIS OF MISPICKS IN WOVEN FABRIC

Images of a woven fabric with missing picks are digitized, and three different image analysis techniques are compared: the Sobel edge operator, the fast Fourier transform, and the discrete wavelet transform. The wavelet transform, used as a multi-resolution spectral filter, is able to give both spectral and frequency information about a fabric. For the samples tested, the wavelet transform can characterize defects due to missing picks and ends faster and more accurately than the other methods.