基于灰度共生矩阵的粗糙表面几何形貌相似性研究

在加工工艺相同的条件下获得待分析表面试样;通过三维表面形貌仪测量获取粗糙表面的几何形貌数据,基于小波变换的方法实现表面形貌的多尺度分解,并滤除基准面;通过灰度共生矩阵方法,计算表面的二阶距/能量、熵、对比度、均匀度四个特征值,进行粗糙表面几何形貌特征值的分析和比较。经小波分析及灰度共生矩阵计算后可知,相同加工参数下的原始表面及各尺度下表面几何形貌的二阶矩/能量、熵、对比度和均匀度都具有很高的相似性,且最大误差<10%。通过试验数据得出:在相同的加工方法及相同加工参数下获得的粗糙表面,其几何形貌在各个尺度之下都具有很高的相似性,为同类型表面的计算机建模及后续分析提供了依据。

Study on Geometry Morphology Similarity of Rough Surface Based on Gray-level Co-occurrence Matrix

The specimens to be analyzed were obtained under the same processing conditions;the surface geometry was measured by three-dimension profilometer and decomposed into multi-scales by wavelet,reference surface filtered.Using gray-level co-occurrence matrix,the four characteristic value of rough surfaces(second moment,entropy,contrast and evenness) were calculated,analyzed and compared.Analyzed by wavelet analysis and gray level co-occurrence matrix,the second moment,entropy,contrast and evenness of the original surface and its components in each scale have high degree of similarity,and the maximum difference is less than 10%.It was found that the geometric morphology of rough surfaces in the same processing condition have high similarity in all scales,which helps to construct surfaces in digital and some further researches.