摘要
表面形貌的分形维数作为表面形貌描述的特征参数与表面形貌灰度图像二值化转换阈值θ密切相关.在灰度图像的二值化转换过程中,合理选取的二值化转换阈值θ*应能够通过从考察对象中所提取的分形维数来较好地表征表面形貌的相关特征.从实例可看到:灰度图像的二值化转换阈值θ存在着极大值θmax和极小值θmin,当θ>θmax或θ<θmin时,从灰度图像二值化转换过程中所获取的计算分形维数的所有信息将会达到饱和状态.基于θmax和θmin,得到了合理选取的二值化转换阈值θ*的经验判据关系.
Fractal dimension, as the characteristic parameter to describe surface topography, relates closely to the grey level threshold θ in the binaryzation of grey image. In the binaryzation process, the rationally selected binaryzation threshold θ^＊ can characterize well the surface topography by way of extracting the fractal dimensions from the objects to investigate. It can be seen from some instances that there are the maximum and minimum thresholds θmax and θmin. When θ〉θmax or θ〈 θmin, all the information obtained from binaryzation to calculate fractal dimensions will be in saturated state. With θmax and θmin taken into account, the empirical criterion θ^＊ can be got as the rationally selected one.
出处
《东北大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2006年第1期57-60,共4页
Journal of Northeastern University(Natural Science)
基金
国家重点基础研究发展规划项目
国家自然科学基金资助项目(50374026)
南京工程学院科研基金资助项目
关键词
表面形貌
图像的二值化
灰度阈值
分形维数
surface topography
image binaryzation
grey level threshold
fractal dimension
