基于冗余提升形态小波的非对称表面特征提取研究

Feature extraction of asymmetric surface topography based on redundancy lifting morphological wavelet

针对提升小波提取非对称表面形貌特征时存在平移变动性和较差的方向选择性,将具有平移不变性和良好的方向选择性的冗余提升形态Haar小波应用到非对称表面的特征提取中。开展了该形态小波对具有显著特征的非对称平顶珩磨的缸套内表面进行特征提取的分析;同时结合仿真的方法,研究了该形态Haar小波对一个模拟的具有一定长度沟槽特征的空间表面进行特征提取的效果。研究结果表明,冗余提升形态Haar小波对非对称表面形貌进行形态特征提取时,在各尺度重构的信号沿特征边缘几乎没有变动且无畸变,可实现准确提取。

Aiming at the shift variation and poor directional selectivity of the feature extraction of the asymmetric surface topography by the lifting wavelet, the redundancy lifting morphological wavelet which is with shift invariance and good directional selectivity was applied to the extracion.Tben the feature extraction of the asymmetric cylinder sleeve of plateau honing surface by the morphological wavelet was analyzed, and the simulation was established to predict the effects of the feature extraction of a simulated space surface with the characteristics of a certain length of trench by the morphological wavelet. The results indicate that, when the redundancy lifting morphological wavelet is carried out to extract the feature of the asymmetric surface topography, the reconstructed signal along characteristic edge is without variation and distortion nearly, and the furture is extracted accurately.