Study on Distilling the Useful Information in Fourier Transform Profilometry

In Fourier transform profilometry, the key problem is how to filter out the fundamental frequency, which concerns the accuracy of the whole system. Four filtering methods are chosen to filter the distortion grating image, and three of them are normal filters, another is wavelet transform method. In simulation experiments, the error distributions of all methods are compared without noise and with noise. The results indicate that: The digital weighted characteristic of hanning filter can restrain the useless signal to some extent; based on time-frequency feature of wavelet transform, the useful frequency component can be distilled while frequency aliasing is restrained.