傅里叶变换轮廓术中基于经验模态分解抑制零频的方法  

Empirical Mode-Decomposition Method for Eliminating Extention of Zero Component in Fourier Transform Profilometry

作　　者：李思坤[1] 陈文静[1] 苏显渝[1] 向立群[1] 

机构地区：[1]四川大学电子信息学院,四川成都610064

出　　处：《光学学报》2009年第3期664-669,共6页Acta Optica Sinica

基　　金：国家自然科学基金(60677028)资助课题

摘　　要：为了消除傅里叶变换轮廓术中零频成分的扩展对测量范围和精度的影响,将经验模态分解方法引入到傅里叶变换轮廓术中,对变形条纹图进行经验模态分解,将条纹图分解为一系列的从高频到低频排列的固有模态函数,达到将高频成分和低频成分相分离的目的,用以消除零频成分,提高测量范围。同采用相移消除零频成分的技术相比,此方法只需要一帧条纹图,测量装置简单、实时性强、计算速度快。文中给出了理论分析和实验验证。In order to avoid frequency aliasing caused by the expansion of zero component in Fourier transform profilometry, empirical mode decomposition is introduced here, by which only one frame deformed fringe pattern is needed. The deformed fringe pattern is discomposed into a series of intrinsic mode functions, which vary form high frequency to low frequency by using empirical mode decomposition. So the high-frequency components are separated from the low ones. The zero component could be eliminated, which leads to the measurement range of FTP to be extended to nearly three times like using the π phase-shifting method. Experiments verify that it is an efficient method to extend the measurement range and keep the measurement speed.

关 键 词：光学测量 傅里叶变换轮廓术 经验模态分解 频谱混叠 

分 类 号：O428.2[理学—声学]