摘要
由于欧几里德几何体系在对复杂形貌度量方面的局限性，传统的图象处理与分析系统对材料表面形貌的测量停留在定性或半定量水平上。近年来，基于分形几何学的若干测量方法如湖岛法（Ｓｌｉｔｌｓｌａｎｄ）、纵剖法（ＦｒａｃｔｕｒｅＰｒｏｆｉｌｅ）等在定量化方面有进展，但它们又存在有损伤探测等弱点。本文提出了基于离散分形布朗增量随机（ＤＦＢＩＲ）场图象模型的无损、自动地求取不规则区域的分形参数的方法。
Because of the limitation of the Euclidean Geometry System on the measurement of complex surface shapes, the measurement of material surface shapes by traditional image processing and analyzing system had been staying on qualitative or half quantitative level all the time. In recent years, several measurement methods based on fractal geometry, such as slit island, fracture profile, have made progress on quantitative analysis. However, both of them have a disadvantage of doing damage to the probed surface. In this paper, a method based on DFBIR image model is presented, which can calculate the fractal parameter of irregular regions automatically and without damage, so that it provides a new quantitative method for the analysis of material images.
出处
《数据采集与处理》
CSCD
1996年第2期103-106,共4页
Journal of Data Acquisition and Processing
基金
高等学校博士学科点专项科研基金
关键词
图象处理
材料图象
分形
材料表面
测量
material science
image processing
fractal image
quantitative analysis
