摘要
为解决图像分析堆积态骨料颗粒群时存在的识别精度问题,应用平板扫描仪,提出了基于灰度形态学重建的粘连颗粒分割方法和基于最小凸多边形边界拟合的骨料图像优化技术.采用图像法分析骨料颗粒群特征参数,计算出图像分析中最小样本颗粒数为720.结果表明:基于灰度形态学重建的骨料颗粒群图像分析方法能解决骨料颗粒粘连时的识别问题,并计算出骨料级配、针度、圆度、球度、棱角性指数、比表面积和针状颗粒含量等颗粒群特征参数.
In order to solve the problem of recognition accuracy of accumulated aggregate particles,a method of image analysis was proposed,using a flatbed scanner.The method included an image segmentation method based on gray morphological reconstruction and an image optimization technique using minimum convex polygons for aggregate boundary fitting.According to the image analysis of the characteristic parameters of aggregate particle group,the minimum number of particles was calculated to be 720 in the image analysis.The results show that the method of image analysis based on gray morphological reconstruction can solve the problem of aggregate particle conglutination,and calculate the characteristic parameters of aggregate particle group such as gradation,elongation,roundness,sphericity,angularity,specific surface area and percentage of elongated particles.
作者
张雄
钟晨
黄廷皓
季涛
ZHANG Xiong;ZHONG Chen;HUANG Tinghao;JI Tao(Key Laboratory of Advanced Civil Engineering Materials of Ministry of Education,Tongji University,Shanghai 201804,China)
出处
《建筑材料学报》
EI
CAS
CSCD
北大核心
2018年第6期886-891,905,共7页
Journal of Building Materials
基金
"十三五"国家重点研发计划项目(2016YFC0700800)
关键词
骨料颗粒群
图像分析
灰度形态学重建
颗粒粘连
aggregate particle group
image analysis
gray morphological reconstruction
aggregate adhesion
作者简介
第一作者:张雄(1956—),男,台湾台北人,同济大学教授,博士生导师,博士.E-mail:zhangxiong@tongji.edu.cn
