Papers by Keyword: Fabric Defects

Abstract: The core of fabric defects detection is the collection and processing of fabrics image. A scheme for fabric defect detection based on cross-entropy is proposed in this paper.The crossentropy value illuminates the information difference between the template image and the realtime image on the average.So can take advantage of cross-entropy criteria to use for defect detection and identification. Results have confirmed the usefulness of this scheme for fabric defect detection.

Abstract: A new technique to inspect fabric defects based on biorthogonal wavelet lifting scheme is presented. In order to get fabric defect detection of rapidity and accuracy, the method is applied to constructs biorthogonal wavelet by lifting scheme and extracts the image characteristic features. The results show that its computing speed is increased by an average of 47% compared with the Mallat algorithm. And it overcomes the lack of space symmetry such as the orthogonal wavelet. The experimental results confirm that it can efficiently inspect four kinds of common fabric defects, warp-lacking, weft-lacking, oil stains, and holes with higher inspection speed.

Abstract: A method to inspect fabric defects based on compactly supported biorthogonal wavelet transform is presented. Firstly, the fabric images are captured by CCD camera. Then fabric defects are detected by means of the strategy of compactly supported biorthogonal wavelet transform. The phase shifts with the orthogonal and the biorthogonal wavelet techniques are compared aiming at the warp-lacking. It is shown that the phase shifts of orthogonal wavelet behave as different degrees, the ones of biorthogonal wavelet are zero. Finally, employing the biorthogonal wavelet method to inspect fabric defects, including warp-lacking, weft-lacking, oil stains, and holes, is given by experiments, in which the results are satisfied.