Textile defect self-adaptive detection method based on spectral analysis

Abstract

The invention relates to the technical field of textiles, in particular to a textile defect self-adaptive detection method based on spectral analysis. The method comprises the following steps of: collecting a surface image of a textile, and obtaining a gray level image and a spectrogram of the surface image; generating different filtering templates for the spectrogram; obtaining a plurality of second grayscale images by using the filtering templates; calculating the offset degree of edge curves in edge images according to the coordinates of pixel points in the edge images of the second grayscale images, and obtaining the density degree of the edge images according to the shortest distance between every two edge curves; obtaining the edge effect of each second grayscale image according to the offset degree, the density degree and the area ratio of the edges; obtaining a weakening effect of each filtering template according to the edge effect; and selecting the second grayscale image corresponding to the filtering template with the maximum weakening effect as a detection image, and determining the defect position of the detection image. According to the method of the embodiment of the invention, adaptive selection of the filtering templates can be completed, and the accuracy of textile defect detection is improved.

Description

technical field [0001] The invention relates to the technical field of textiles, in particular to an adaptive detection method for textile defects based on frequency spectrum analysis. Background technique [0002] In the production process of textiles, it is necessary to use the cloth inspection machine to detect the defects of the final textile products. When many cloth inspection machines detect defects, operators need to observe visually under sufficient light sources to find defects such as surface defects and color differences. The machine automatically completes the length recording and package finishing work. In this case, the detection speed of textile defects is low, and the subjectivity of the inspectors has a great influence on the detection results, and it is easy to miss inspection; the cloth inspection machine with good performance is equipped with electronic defect inspection The device uses image segmentation or neural network to identify defect parts, and t...

AI Technical Summary

Problems solved by technology

[0002] In the production process of textiles, it is necessary to use the cloth inspection machine to detect the defects of the final textile products. When many cloth inspection machines detect defects, operators need to observe visually under sufficient light sources to find defects such as surface defects and color differences. The machine automatically completes the length recording and package finishing work. In this case, the detection speed of textile defects is low, and the subjectivity of the inspectors has a great influence on the detection results, and it is easy to miss inspection; the cloth inspection machine with good performance is equipped with electronic defect inspection device, through image segmentation or using neural network to identify defect parts, and textiles often have a lot of texture background features, which may lead to inaccurate defect detection results

Images