Crosstalk Suppression for Color Phase-Shifting Profilometry Based on Chord Distribution Equalization

Color phase-shifting profilometry encodes the gray fringe pattern of the conventional three-step phase-shifting algorithm into a single-color image as a means of greatly increasing the measurement speed. However, the color crosstalk problem induces phase errors, which affect the measuring accuracy. In this study, a color crosstalk suppression method based on the equalization of the wrapped phase chord distribution is proposed. By using the wrapped phase value as the polar angle, a chord distribution diagram in the polar coordinate can be obtained. The diagram can be regarded as the distribution of the wrapped phase. Then, the polar angle value is used to equalize the chord distribution. After equalization, a uniform chord distribution diagram is retrieved, and it can be used to recover a new wrapped phase. Nonlinear errors caused by the color crosstalk can be suppressed in a manner that corrected the wrapped phase. This method is a completely software-based method for eliminating color crosstalk, which is more applicable and flexible than conventional methods requiring precalibration, the construction of a lookup table of phase errors, or additional patterns. The results of simulations and real experiments indicate the ability of the proposed method to effectively correct the nonlinear errors caused by the color crosstalk.