Automated detection of the first order frequency in Fourier transform profilometry

Abstract

We determined the position of the first order frequency of a vertical sinusoidal fringe pattern in Fourier Transform Profilometry (FTP) using a proposed fringe counting algorithm. It was observed that the position varies proportionally with the frequency of the fringe pattern. With this observation we were able to show that the pixel position of the first order frequency is equal to the fringe count. A fringe counting algorithm is then proposed to determine the number of fringes. Results show that the proposed algorithm was able to obtain accurate fringe counts of the patterns with different frequencies. With this approach automated detection of the first order frequency is achieved which helps in speeding up the algorithm for FTP. In the case where the frequency of the projected fringe pattern is changed. The proposed method will be able to automatically determine the parameters needed for FTP.