Electronic cloth defect segmentation method

Abstract

The invention discloses an electronic cloth defect segmentation method. The method comprises the following specific steps: firstly, converting a to-be-detected electronic cloth image into a gray levelimage; secondly, carrying out ButterWorth filtering processing on the gray level image; clustering the filtered images by using a K-means algorithm; and finally, marking the pixel points of the clustered image according to the clustered category, and carrying out binarization processing according to the marking result, so that fabric defects can be segmented. According to the method, the ButterWorth low-pass filter is used for processing the electronic cloth image, detail information of defects can be reserved to the maximum extent while background textures are restrained, and the contrast ratio of the background and defect areas is enhanced. Meanwhile, the filtered images are clustered through the K-means algorithm, finally, the electronic cloth images can be clustered into different classes, then binary segmentation is conducted on the images according to the marking result, and accurate positioning and accurate segmentation of defects of the electronic cloth are achieved.

Description

technical field [0001] The invention belongs to the technical field of fabric defect detection in the textile industry, and in particular relates to a method for segmenting electronic cloth defects. Background technique [0002] Glass fiber electronic cloth, also known as electronic cloth, is an industrial product woven from glass fiber yarn. As a reinforcing material, it is widely used in aerospace, mechanical parts and other fields. In the process of industrial production, defects are an important factor affecting its price and grade evaluation, and it is also a key link in product quality control. At present, most electronic cloth manufacturers at home and abroad mainly rely on manual inspection of electronic cloth defects. Manual detection is greatly affected by subjective factors, such as long-term visual fatigue, working environment, etc. The defect detection rate is only about 70%, and there are problems such as slow detection speed, low accuracy rate, and high misse...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a method for segmenting electronic cloth defects, which solves the problem of inaccurate positioning and inaccurate segmentation of defects on electronic cloth by detection methods in the prior art

Images