Comparison of real-time phase-reconstruction methods in temporal speckle-pattern interferometry

Three real-time methods for object-phase recovery are implemented and compared in temporal speckle-pattern interferometry. Empirical mode and intrinsic time-scale decompositions are used and compared as real-time nonstationary and nonlinear filtering techniques for the extraction of the spatio-temporal evolution of the object phase. The proposed real-time methods avoid the application of the Hilbert transform and improve the accuracy of the measurement by filtering under-modulated pixels using Delaunay triangulation. The performance of the proposed methods is evaluated by comparing phase-recovery accuracy and computation time by means of numerical simulations and experimental data obtained from common and simultaneous π/2 phase-shifting heterodyne interferometry.