A Phase Optimization Method Based on Fast Iteration and Boundary Guidance

To address the phase error problem of the objects with large gradients and low reflectivity in phase shifting (PS) profilometry (PSP), we put forward a novel strategy in which the invalid points are first identified and removed automatically, and then the fringe orders of the residual points are corrected. Aiming at the invalid points, an invalid point location and removal method based on fast iteration (FI-IPLR) is proposed. Compared with the manual threshold selection method, the FI-IPLR method can quickly and adaptively determine the threshold to remove invalid points. For the residual points, we first analyze the mechanism of phase error generation and then present a closed-loop strategy that integrates error point identification, fringe order modification (FOM), and verification methods. To avoid the normal points from being miscorrected, we first present an error point location procedure based on multiindicator union to eliminate the pseudo-error points and identify the real error points. After locating the real error points, a FOM method based on boundary guidance (FOM-BG) is proposed to modify their fringe orders. Then, a fringe order verification and rectification method based on monotonicity is used to check and repair the modified fringe order. Finally, we use the modified fringe order to calculate an optimized absolute phase map. The experimental results demonstrate that the proposed method performs better than other methods, especially for the large gradient, discontinuous, and low reflection objects.