Multi-frequency phase unwrapping algorithms (MFPUAs) are among the most robust and efficient phase unwrapping methods, and hence widely adopted in various optical measurement techniques, such as fringe projection profilometry, interferometry, and so on. However, if they are not well designed, or the measurement condition is poor, fringe order error (FOE) will be triggered during the phase unwrapping, which will result in phase unwrapping error. This paper is for coping with FOE without consuming additional fringe patterns for phase unwrapping. Firstly, a parameter called fringe order inaccuracy (FOI) which is applicable to the representative MFPUAs is defined. Secondly, FOI is employed to help quantify the possibility of the occurrence of FOE. Lastly, a fast FOE correction method is proposed with aid of FOI. Experiments validate that the proposed method contributes to suppressing FOE while keeps the efficiency of MFPUA.
Phase unwrapping is a significant procedure that has raised a great interest in many coherent imaging systems. What we believe to be a new phase unwrapping algorithm, is described and tested. The method starts from the fact that 2D wrapped phase distribution can be regarded as a response to two orthogonal 1D direction excitation signals. This suggests a cepstrum analysis to be implemented in the phase unwrapping problem. Experimental results both from the fringe projection profilometry and DENSE MRI also confirmed the validity of our approach. In fact, this proposed method is possible to attain a fast and practical phase unwrapping solution with enhanced noise robustness.
A phase unwrapping algorithm specially designed for phase-shifting fringe projection profilometry (FPP) is proposed. This algorithm is based on a principle that combines a refined double-frequency fringe projection algorithm and a fringe background based quality guided phase unwrapping algorithm (QGPUA). The phase demodulated from high-frequency fringe images are partially unwrapped by that from the low-frequency ones. This would help to avoid phase order ambiguity in the eventual phase unwrapping while guarantee the phase resolution. The fringe background based QGPUA is adopted to completely unwrap the partially unwrapped phase. The quality map utilized during the phase unwrapping is generated from the fringe background and divided into different subregions. Further unwrapping for the partially unwrapped phase goes on in the order from phase in the highest quality subregion to that in the lowest one gradually. Partially unwrapped phase in each subregion is unwrapped with flood-fill algorithm. Therefore the phase unwrapping speed can be improved. The proposed method can make the phase unwrapping for phase-shifting FPP more robust and faster. Experiment result confirms the effectiveness of the proposed method.