西部皮革2024,Vol.46Issue(14) :12-14.DOI:10.20143/j.1671-1602.2024.14.012

基于相对总变差与自适应阈值分割的织物疵点检测算法

Algorithm for Fabric Defect Detection based on Relative Total Variation and Adaptive Threshold Segmentation

王春妍 李丽瑶 白少云
西部皮革2024,Vol.46Issue(14) :12-14.DOI:10.20143/j.1671-1602.2024.14.012

基于相对总变差与自适应阈值分割的织物疵点检测算法

Algorithm for Fabric Defect Detection based on Relative Total Variation and Adaptive Threshold Segmentation

王春妍 1李丽瑶 1白少云1
扫码查看

作者信息

  • 1. 山西电子科技学院智能制造产业学院,山西临汾 041000
  • 折叠

摘要

在纺织领域,织物疵点智能检测一直是亟须解决的难题之一,为了对织物疵点进行自动检测,提出基于相对总变差模型与自适应阈值分割的织物疵点检测算法.该方法选用中值滤波降噪,然后通过相对总变差模型增强织物疵点,抑制织物背景纹理,接着对疵点增强图像的灰度直方图进行分析,利用迭代的自适应阈值分割疵点,得到二值图像.经对多种类型疵点图像进行测试分析,此算法能增强疵点并有效分割,有较好的检测结果.

Abstract

In the textile industry,intelligent detection of fabric defects remains a formidable challenge.To automate the detection of fabric defects,we propose an algorithm based on the relative total variation model and adaptive threshold segmentation.This approach utilizes median filtering for noise reduction,enhances fabric defects using the relative total variation model to suppress background textures,and analyzes the grayscale histogram of the enhanced defect image.Through iterative adaptive threshold segmentation,a binary image of defects is obtained.Testing on various types of defect images demonstrates that this algorithm effectively enhances and segments defects,yielding promising detection outcomes.

关键词

疵点检测/中值滤波/相对总变差模型/自适应阈值分割

Key words

defect detection/median filtering/relative total variation model/adaptive threshold segmentation

引用本文复制引用

基金项目

山西省高等学校科技创新项目(2023L466)

出版年

2024
西部皮革
四川省皮革行业协会 四川省皮革学会 四川省皮革研究所

西部皮革

影响因子:0.293
ISSN:1671-1602
段落导航相关论文