基于纹理提取的针织物密度检测  

Knitted Fabric Density Detection Based on Texture Extraction

作　　者：于东洋 柯薇[1] 刘威 邓中民[1] Yu Dongyang;Ke Wei;Liu Wei;Deng Zhongmin(State Key Laboratory of New Textile Materials and Advanced Processing Technology,Wuhan Textile University,Wuhan,Hubei 430200,China)

机构地区：[1]武汉纺织大学省部共建纺织新材料与先进加工技术国家重点试验室,湖北武汉430200

出　　处：《针织工业》2022年第10期19-22,共4页Knitting Industries

基　　金：湖北省技术创新专项(2019AAA005)

摘　　要：为提高针织物密度检测准确度和检测效率,在针织物图像基础上,通过增强图像对比度并运用理想高通滤波实现织物高频增强提取针织物圈柱图像,运用形态学运算处理针织物圈柱图像,检测得到图像中圈柱数目。将圈柱图像细化处理获取圈柱骨架,针对骨架特点提取纵行纹理,进而检测针织物纵纱数量,得到横向线圈数,根据总的圈柱数目和纵纱数目得到纵向圈柱数,进而计算针织物线圈密度,实现针织物密度的自动检测。试验结果表明:该方法可以快速、准确地检测纬编针织物密度。In order to improve the fabric density detection accuracy and efficiency,on the basis of knitted fabric image,by enhancing the image contrast and high-pass filter to realize ideal fabric knitted fabric side limb image extracting high-frequency enhancement,using morphological image processing knitted fabric side limb,detected by the image circle number column,column image processing to obtain the side limb skeleton,longitudinal texture analysing according to the characteristics of the skeleton extracted,then testing longitudinal yarn knitted loop quantity,horizontal loop number,according to the general side limb number and longitudinal yarn number,knitted fabric stitch density are calculated,and realize the automatic detection of fabric density.The experimental results show that the method can quickly and accurately detect the density of weft knitted fabric.

关 键 词：针织物图像 针织物检测 针织物密度 理想高通滤波 细化处理 自动检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TS187[自动化与计算机技术—计算机科学与技术]