Expression And Fractal Analysis Of Morphology Structures Of Crepe Fabric Surface Texture

Crepe fabric is a fabric which though weave or different technological conditions. Crepe fabric has different appearance winkle, a soft sheen and handle. Crepe fabric has a smoothness, roughness, irregularly appearance texture. Crepe effect is a kind of reflecting of winkle features in people's mind. Crepe effect is the major style of crepe fabric. So far, experts have made effective research on crepe effect mechanism. However, we have many difficulties in expounding crepe fabric winkle power,controlling crepe effect level and evaluating crepe effect. All of these, evaluating crepe effect is the most difficult one, which has no unified standard. Therefore, if a direct structural parameter for crepe fabric surface texture can be found, it well be very useful for crepe fabric production and evaluation. The relationship among structural parameter, winkle degree and visual effects can be got.Matlab is a highly integrated system which contained scientific calculation, image processing and speech processing. In this paper, five kinds of crepe fabrics with different winkle degree are given. Matlab is used to deal with appearance wrinkle image by eliminate correction, increase contrast and image filter, then binary image can be got based on morphological image processing. Two-dimensional Fourier transform profilometry without frequency-moving for data acquisition of three-dimensional transformation of fabric appearance is proposed. In this method, Crossed-Optical-Axis Geometry and phase unwrapping technique based on digital weighted filter and reliability mask are employed. Crepe fabrics surface section contour image based on digital weighted filter is employed. Parameters are extracted from the binary image and the reconstructed 3-D surface shape, such as creping degree, stripe repeating frequency, duplicating degree of stripe shape, surface areas, coarse degree, twist degree, kurtosis, average offset and sharpness. They are found to be effective of express the irregular morphologic structures of crepe fabric surface texture. The calculation results show that the correlation coefficients between these feature parameters and winkle grade are better.Fractal theory, which is a nonlinear theory developed with in recent thirty years, is generally used in nature. Fractal theory provides an great mathematical method researching, in irregular complex objects in nature. Bashed on computer image analysis, fractal theory is used to analysis the binary image of crepe fabrics, restored surface and fabric surface section contour image bashed on Fourier transform profilometry of crepe fabrics. The experimental results show that the correlation coefficients between these fractal dimensions and winkle degree are batter, which means all the fractal dimensions can be further used to evaluate the winkle degree of crepe fabrics.The result shows that latent parameters and fractal dimensions have close relationship with winkle degree. Crepe fabrics with similar winkle degree or coarse degree, the latent parameters and the fractal dimensions are similar, too. There is obvious difference in latent parameters and fractal dimensions between fabrics with different winkle degree. Accordingly, the latent parameters and the fractal dimensions are used to describe winkle degree.