摘要
根据毛粘混纺纱横截面切片图像,应用Matlab对纱线组分进行有效的分割,讨论了在分割图像处理中关键参数选取的客观性。包括:根据切片灰度图像盲孔特征,应用最大类别方差法(Otsu法)确定二值化域值;应用regionprops函数的Solidity属性,通过区域面积随Solidity阈值分布频率图,确定区分纱线不同组分的Solidity域值。这2个关键参数确定方法比较客观,实践中获得了符合实际的图像预处理和分割效果。
This article divided up the component of yarn by Matlab according to cross section image of blended wool /viscose yarns,discussed that objectivity of key parameter selected during dealt with the image division.It included that applied the method of Otsu to confirmed the binarization value according to the character of slice gray image blind via,applied the solidity attribution of regionprops function,and according as the distributing frequency chart that area of zone follow the solidity threshold value to confirmed the distinguished solidity value of yarn which different component.The confirm method of those two key parameters were objective,achieved the effect of image pretreatment and division according with the true during the practice.
出处
《毛纺科技》
CAS
北大核心
2010年第6期59-62,共4页
Wool Textile Journal
关键词
MATLAB应用
图像处理
纱线检测
组分分析
Matlab application
image treatment
yarn inspection
analysis of component