计算机技术与发展2011，Vol.21Issue(11)：120-123,4.
焦炭显微图像的轮廓提取新方法研究
A Novel Contour Extraction Method for Coke Microscopic Image
摘要
Abstract
The extraction of regional boundary of coke microscopic image is the key to classify and recognize different optical components. Because the optical components in different coke polarized showed diversity, and the complexity of the image itself, the existing methods extraction boundary meats great difficulties. In this paper,use one method based on boundary weighted meanshift clustering, it better keeps edge information. Then, adopt estimate threshold and polygonal approximation to get the continuous boundary of the image. And experiments have shown that this method can denoise with high efficiency, and can extract the boundaries of different compounds of image accurately. Make it possible to classify certain categories of coke microscopic image using the contour feature information.关键词
焦炭光学组织/边界加权/均值偏移/双阈值法Key words
coke optical organization/boundary weighted/meanshift/estimate threshold分类
信息技术与安全科学引用本文复制引用
缪家龙,王培珍,孙文龙,余坚毅..焦炭显微图像的轮廓提取新方法研究[J].计算机技术与发展,2011,21(11):120-123,4.基金项目
国家自然科学基金项目(50874001) （50874001）
