摘要
针对扫描电子显微镜图像中粉煤灰区域的提取问题,提出了基于区域生长的珠状颗粒区域提取方法。该方法自动选取种子区域,在边界约束下按灰度相似度生长,生长区域结合形状特征提取目标。提取结果比对人工提取区域,提取误差用提取面积的错分率和漏分率之和表示。实验的最小提取误差是6.8%,平均提取误差是8%,自动提取60张SEM图像的时间小于10 min。提取时间和误差均可达到估算粉煤灰含量的要求。
An unsupervised extraction method was proposed in order to extract bead-like particle regions of fly ash from scanning electron microscope image,which was based on region growing with gray similarity bounded by gradient and shape.The process was automatic,including seeds selection,regions growing and shape distinguishing.The experimental error was measured by the acreage probability of missing segmentation and false segmentation.The minimum error rate of the experimental results was 6.8 percent,and the average error rate was 8 percent.The time of extraction from 60 SEM images was within 10 minutes.The method is effective for the content estimation of fly ash in the material.
出处
《计算机应用》
CSCD
北大核心
2012年第6期1570-1573,共4页
journal of Computer Applications
基金
国家自然科学基金资助项目(61070227)
国家自然科学基金-广东联合基金重点项目(U1135003)
973计划项目子课题(2009CB62310504)
关键词
扫描电子显微镜图像
粉煤灰
梯度
区域生长
自动提取
Scanning Electron Microscope(SEM) image
fly ash
gradient
region growing
automatic extraction