摘要
本文在使用扫描电子显微镜(SEM)获得软土微观图像基础上,将图像进行灰度化处理和二值化处理,使用Canny边缘检测技术得到了软土微观图像中颗粒的分布情况与土的面孔隙比的大小,进而获得了偏心率、扁圆度、圆形度、离散指数、内切圆半径等土颗粒形状特征参数,进行了这些形状特征参数之间的相关性分析和因子分析。结果表明,偏心率和各向异性率等参数之间具有很好的相关性;周长和离散指数等参数之间具有较好的相关性;而各向异性率和面积等参数之间则不具有相关性;所有特征参数可以由区域因子、周长因子和弦因子等3个主因子来反映。
Soil microstructure has a significant effect on the engineering properties of the soil.In the current study,the images of a soft soil are obtained using the Scanning Electron Microscopy(SEM) technique.After converting the original images into the grayscale and binary ones,the Canny edge detection technique is utilized to detect the particle edges and compute the planar porosity ratio of the soil and the characteristic shape parameters of the particle in the SEM images,including eccentricity,flat circular degree,circular degree,discrete index,radius of inscribe circle.The analysis of correlation and factor analysis between these shape parameters are performed.The results show that some parameters,such as eccentricity and anisotropic rate,has a close correlation;other parameters,such as perimeter and discrete index,has a good correlation;and some parameters,such as anisotropic rate and area,has no correlation.All of the characteristic parameters are reflected by 3main factors,including the region,perimeter and chord parameters.
出处
《水文地质工程地质》
CAS
CSCD
北大核心
2013年第3期69-73,109,共6页
Hydrogeology & Engineering Geology
基金
国家自然科学基金项目(40972191)
关键词
边缘检测
形状特征参数
相关性
因子分析
edge detection
characteristic shape parameters
correlation
factor analysis