Automatic Recognition Parameters Of Fabric Construction By Adaptive Wavelets

The mechanical properties of a woven fabric depend not only on those of the yarns constituting the fabric, but also on the structural properties of the fabric itself, such as weave pattern, yarn number, fabric density. At present, inspection of these fabric characteristics still relies on manual operations, which are time-consuming and easily tire an operator's eyes. Thus, it is highly desirable to develop image processing and automatic analysis techniques to identify fabric characteristics.The automatic recognition parameters of fabric construction are studied in this paper. In chapter 1, different image analysis methods for fabric construction recognizing are introduced and the adaptive wavelet analysis method is selected in this paper. The basic theory of wavelet transform and the constructing method of wavelet filters are explained in the second chapter. The experiment equipment and image pretreatmentincluding histogram equalization, Wiener filter etc are introduced in detail in the third chapter. The process of automatic recognition by adaptive wavelet transform and autocorrelation are designed in the forth chapter. In the fifth chapter the adoptive method are discussed by fabric construction, the color of the fabric, the magnifying and different recognizing area. In the sixth chapter, the conclusion of this paper is concluded, Feather more, the next step of this project is showed.In this paper, we attempt to develop an efficient system based on the analysis method of adaptive wavelet which is an important method of image processing. We capture the fabric image with reflective light, process it by digital image analysis technology, analyze the image to obtain one repeat of the weave, and determine the interlacing states at the crossover points. Then the fabric pattern will be automatically recognized. At the same time, we can measure the weaving density in the image processing. The results demonstrate that three fundamental weave types can be classified accurately, and structural parameters such as yarn spacing can also be obtained.