Two-step phase extraction and random phase shift estimation in phase-shifting profilometry based on least-squared optical flow method

Phase shifting profilometry normally requires at least three-frame fringe patterns to extract the phase of objects. In the dynamic measurement field, it is desirable to reduce the number of fringe patterns and eliminate the phase-shifting error caused by the moving of the test object between fringe patterns. In the field of interferometry, the optical flow technique is an outstanding two-step phase reconstruction method. However, the accuracy of the optical flow (OF) algorithm is limited by the residual background intensity and approximation operation as well as noise. In this paper, a phase reconstruction method based on the optical flow algorithm and two-frame least square algorithm (OF&LSA) is proposed to alleviate the phase error caused by the above problems. The proposed method is employed in dynamic phase-shifting profilometry to achieve 2 -frame phase reconstruction. The simulation and experiment results verify that the proposed OF&LSA improves the phase accuracy of OF algorithm significantly, and it also performs better than the 3-step phase shifting method in our dynamic measurement experiment.