A threshold segmentation method for a yarn hairiness image

Abstract

The invention discloses a threshold segmentation method for a yarn hairiness image. A linear region threshold segmentation method is used for carrying out yarn and background segmentation on a preprocessed image, and the method mainly comprises the steps of calculating an inter-class variance with each column as a whole, segmenting the inter-class variance with an optimal threshold, combining theimages obtained after threshold segmentation of each column, and outputting an optimal threshold segmented image. Compared with an existing inter-class variance threshold segmentation method, the yarncan be accurately separated from the background, the yarn image information loss is reduced, and meanwhile the follow-up yarn extraction processing is facilitated.

Description

technical field [0001] The invention relates to the technical field of textiles, in particular to a method for threshold value segmentation of yarn hairiness images. Background technique [0002] In recent years, with the technological innovation and progress of the textile industry, higher requirements have been put forward for the quality inspection of textiles. Factors affecting the appearance and quality of textiles appear at all stages of the textile production process, the most common of which is the quality of the yarn, and the most important feature of the quality of the yarn is the appearance of the yarn hairiness, which is a measure of One of the important indicators of yarn quality, it has a significant impact on yarn performance, quality and subsequent processing. Yarn hairiness has complex appearance characteristics, and the outer fiber protruding from the main body of the yarn is the main part of the hairiness. Therefore, the end hairiness or ring hairiness pr...

AI Technical Summary

Problems solved by technology

[0010] Using the OTSU inter-class variance threshold segmentation method will affect the segmentation results of the entire image due to the background noise of a certain part of the image, such as Figure 1-2 and Figure 3-4 The filtered image is subjected to OTSU threshold segmentation, the result is as follows Figure 5-6 and Figur

Images