双密度小波在表面形貌信号分离中的应用

小波的时频分析特性，使其在表面功能评定中克服了传统滤波方法的缺陷，可以合理实用地评定表面，但现有的小波模型中存在着滤波结果与测量原始数据的起始采样点位置有关的缺点，为此提出了一种双密度小波变换的表面形貌分离的新方法，用于工程表面的分析，实现了对表面特征平移不变的提取，提高了分离粗糙度、波度等频率成分的准确性。

Surface topography signal separation with double density wavelet transform

With the characteristic of its time-frequency analysis, wavelet can overcome the defect of the traditional filtering methods and appropriately evaluate the surface topography. This paper proposed a novel signal separation method based on double density wavelet transform (DDWT) for engineering surface analysis. The method provides near shift-invariance in surface texture extraction and improves the separation accuracy of roughness, waviness and other frequency component.