图像分维对腐蚀坑分布特征的表征

CHARACTERIZATION FOR CORROSION PIT DISTRIBUTION BY USING FRACTAL DIMENSION OF IMAGE

采用图像分维探讨了腐蚀图像特征与表面腐蚀坑分布特征的关系.统计碳钢土壤腐蚀试片表面测量数据表明,蚀坑直径和深度分布在一定尺度范围内遵循幂函数方程,即蚀坑直径或深度与相应出现概率(个数)在双对数坐标存在线性区,其线性区斜率代表坑直径或深度的分布分维,反映腐蚀表面的粗糙程度.同时,根据盒计数法原理提出计算图像二维和三维分维的方法.大量实验数据证实,腐蚀图像二维和三维分维分别与坑直径和坑深度分布分维线性相关,并在数值上近似相等.这为根据腐蚀图像特征确定腐蚀表面坑分布规律提供了简便快捷的方法.

The relativity between image feature and pit distribution characteristic of corroded surface is studied. According to measure data of carbon steel samples corroded in soil, it can be sure that the distributions of corrosion pit diameter and depth obey exponential function equation in certain region. In other words, there exists a good linear section between corrosion pit diameter or depth and corresponding appearance probability in a double logarithmic coordinate system, which linear slope is fractal dimension （FD） for pit diameter or depth distribution denoting the roughness of corroded surface. Simultaneously a new approach is proposed based on box - counting method to estimate the fractal dimensions, D2d and D3d of corrosion images. Use a great number test data to prove a good linear relation between image D2d and D3d respectively with FD for pit diameter distribution and pit depth distribution. They are approximating equality in value. This result provides a simple and rapid method to research the rule of pit distribution on real corroded surface according to corrosion image feature.