Studies On Recognition Of Fabric Weave Parameters Based On Wavelet Analyse

With the popularization of computer digital image processing technology, in the last more than ten years, the textile field, like other academic fields, has started to make use of the image processing technique to analyse and examine textile, thus create the relation between the microstructure and macroscopic property.Fabric density is an important index in the National Standards of Calico. In Estimation of Fabric Density GB/T4668-1995, three methods are recommended: a) fabric decomposition;b) fabric analyzing lens;c) portable fabric density estimating lens. Of the three mentioned methods, the fabric decomposition method enjoys a very good accuracy and prevails in the textile industry. However, this method involves a heavy workload on the part of the checkers who must concentrate on the counting and any miscounting will break the estimation process and the checkers have to start counting again, which is time and energy consuming, and that will damage the fabric samples.This paper aimed at the existent limitation of current fabric examination, has carried through thorough research, applied computer digital image processing technology to carry on automatic recognition, and developed automatic recognition software of fabric weave parameters, which advanced the efficiency and the accuracy of the fabric design. The paper based on the following three assumptions:1) Weft and warp arrange equably and tightly.2) The change in weft's brightness reflects weft and warp points;3) The position of weft in fabric is decided by peek and valuebrightness projection curve in the direction of weft;The original fabric images (the reflection image) were obtained byusing scanner. First, put prepare process on original fabric image, whichconvenient for the follow step, then decompose the processed images withWavelet decomposition, the decomposed images include longitude andlatitude fabric information, adopting the method of data average to get rid of the impact caused by fabric deformation and scan image deviation (mainly make use of the projection information of longitude and latitude of the fabric), second, adopt Fourier transform to calculate the period of fabric yarn, then get the yarn density of longitude and latitude;as for the diameter of the yarn, we dichotomize the decomposed images and calculate an average diameter of yarn;with the density of longitude and latitude and diameter of yarn, we get the location of the longitude and latitude yarn at last.The system mainly takes Matlab as platen to carry on the image processing, and combining with VC++6. 0 as develop tool, establish the fabric automatic recognition system consumer interface.