摘要
工程应用中采用工业电子计算机断层扫描(computed tomography,CT)技术对金属材料内部缺陷进行检测,基于等值面分割法对缺陷进行尺寸测量,然而该方法存在小缺陷定量精度低、稳定性差等问题。对此提出了一种利用CT图像局部区域灰度变异系数表征小缺陷尺寸的方法。首先,设计制作不同穿透厚度、缺陷直径(通孔)的对比试块,对其进行CT检测获得模拟小缺陷断层图像,分析图像局部灰度特征,确定变异系数与小缺陷尺寸的对应关系,建立缺陷尺寸定量方法,并与半高宽法、等值面分割法相比较。研究结果表明:变异系数与小缺陷当量尺寸的相关系数稳定在0.99以上,基于变异系数定量方法的精度及稳定性高、操作简便、实用性强,是作为小缺陷定量检测的有效途径之一。
In engineering applications,industrial computed tomography(CT)technology was usually used to detect small defects inside metal materials through isosurface segmentation.However,the small defect quantification method had problems such as low accuracy and poor stability.It was proposed that using the differences in grayscale variation coefficients of small defect in CT images to characterize defect sizes for higher measurement accuracy.Firstly,the contrast blocks with different penetration thickness and defect diameter(through-hole)were designed and manufactured,and then CT inspection was carried out to obtain the simulated small defect fault image.The local gray characteristics of the image were analyzed to determine the corresponding relationship between the coefficient of variation and the size of small defects,and the quantitative method of defect size was established,which was compared with half height width method and isosurface segmentation method.The research results show that the correlation coefficient between the coefficient of variation and the area of small defect is stable above 0.99.The quantitative method based on the coefficient of variation is more accurate and stable than the full width at half maxima method(isosurface segmentation method).This method is easy to operate,and has high practicability.It is one of the effective ways to detect small defects.
作者
齐子诚
倪培君
张维国
姜伟
裘信国
QI Zi-cheng;NI Pei-jun;ZHANG Wei-guo;JIANG Wei;QIU Xin-guo(The Ningbo Branch of Ordnance Science Institute of China,Ningbo 315103,China;College of Mechanical Engineering,Zhejiang University of Technology,Hangzhou 310014,China)
出处
《科学技术与工程》
北大核心
2021年第3期958-964,共7页
Science Technology and Engineering
基金
国家自然科学基金(61701446)
领域基金项目(6140923030503)
浙江省基础公益项目(LGG20F010003)
宁波市科技服务示范项目(2019F1036)。
关键词
工业CT
金属材料
小缺陷
变异系数
定量方法
industrial CT
metallic materials
small defects
coefficient of variation
quantitative method
