Automatic measuring method of fabric image texture cycle

Abstract

The invention relates to an automatic measuring method of a fabric image texture cycle. The method comprises the steps that a two-dimensional gray-level fabric image is expanded to one-dimensional row vectors and one-dimensional column vectors in the row mode and the column mode respectively, the one-dimensional row vectors and the one-dimensional column vectors are subjected to a rapid one-dimensional Fourier transformation to obtain a corresponding continuous spectrum, a principle period d1 and a secondary period d2, corresponding to the maximum peak point and the secondary peak point of the continuous spectrum respectively, are extracted, the distance matching function value DMF1 and DMF2 of the d1 and the d2 are calculated respectively, and making the peak point ratio K is equal to d2 / d1, when the value of K is lower than a threshold value, the ultimate period is d1, or the ultimate period is min (DMF1, DMF2). According to the method, on the basis of the Fourier transformation, the fabric texture period is automatically measured combining the distance matching function, and the method has better stability, higher calculation efficiency, and better measuring effect on a small period texture which is hard to be recognized by most measuring methods, and can be applied to the automatic measuring field of fabric longitude and latitude texture periods.

Description

technical field [0001] The invention belongs to the field of image analysis and processing, and relates to an automatic measurement method for fabric image texture period based on the combination of frequency domain analysis and distance matching function. Background technique [0002] In the process of textile processing and testing, a large number of problems involve the surface texture and surface shape of fabrics, and the period is one of the important visual features of texture images. At present, the existing methods for measuring the period of texture images mainly include: frequency domain-based Fourier method, autocorrelation function method, gray level co-occurrence matrix method and Renyi generalized entropy method. [0003] In 1999, Xin Binjie and others proposed to determine the texture period by locating the main frequency in the Fourier frequency domain. Although this method can accurately obtain the frequency components contained in the image, it is greatly a...

AI Technical Summary

Problems solved by technology

In 2012, Asha et al. proposed the accumulative DMF texture period extraction method, combined with the first two differences, to realize the fully automatic extraction of texture row and column periods, but this method is not suitable for textures with noise or sub-periods. Poor stability

Grigorescu et al. proposed a texture period extraction method based on Renyi generalized entropy in 2003. This method counts the Renyi entropy of the texture structure between sub-windows of different sizes. When it reaches the minimum value, it is the period of the texture or its integer. times, but this method is greatly affected by the random irregularity of the texture and needs to manually select the local minimum point

Images