Abstract:Phase is an inherent characteristic of any wave field. Statistics show that greater than 25% of the information is encoded in the amplitude term and 75% of the information is in the phase term. The technique of phase retrieval means acquire phase by computation using magnitude measurements and provides data information for holography display, 3D field reconstruction, X-ray crystallography, diffraction imaging, astronomical imaging and many other applications. Mathematically, solving phase retrieval problem is an inverse problem taking the physical and computation constraints. Some recent algorithms use the principle of compressive sensing, such as PhaseLift, PhaseCut and compressive phase retrieval etc. they formulate phase retrieval problems as one of finding the rank-one solution to a system of linear matrix equations and make the overall algorithm a convex program over n × n matrices. However, by "lifting" a vector problem to a matrix one, these methods lead to a much higher computational cost as a result. Furthermore, they only use intensity measurements but few physical constraints. In the paper, a new algorithm is proposed that combines above convex optimization methods with a well known iterative Fourier transform algorithm (IFTA). The IFTA iterates between the object domain and spectral domain to reinforce the physical information and reaches convergence quickly which has been proved in many applications such as compute-generated-hologram (CGH). Herein the output phase of the IFTA is treated as the initial guess of convex optimization methods, and then the reconstructed phase is numerically computed by using modified TFOCS. Simulation results show that the combined algorithm increases the likelihood of successful recovery as well as improves the precision of solution.

Phase Recovery Technology of a Dual-Frame Phase-Shifting Interferogram Based on First-Order Norm Vector Normalization

Selection of Phase-Recovery Algorithms for Fringe Processing in Optical Measurement of Micro-Surface

Single-frame Noisy Interferogram Phase Retrieval Using an End-to-end Deep Learning Network with Physical Information Constraints

1-Norm Character of Phase Shifting Interferograms and Its Application in Phase Shift Extraction

Dual-frame Demodulation Based on an Advanced Vector Normalization Technology

Dual-wavelength Efficient Phase Imaging Method Based on Convolutional Neural Networks

Phase Retrieval Approach Based on the Normalized Difference Maps Induced by Three Interferograms with Unknown Phase Shifts

An advanced phase retrieval algorithm in N-step phase-shifting interferometry with unknown phase shifts

Accurate phase retrieval algorithm based on normalized vector of three interferograms

Phase-shifting Interferometry by a Similar Matrix 1-Norm Algorithm

Rapid phase shifts extraction and phase retrieval by performing 1-norm operation for phase-shifting interferograms

Two-frame phase extraction technology based on vector normalization with single filter

A precise three-step phase retrieval algorithm suitable for the interferograms with less than one fringes numbers

Two-step Orthogonalization Phase Demodulation Method Based on a Single Differential Interferogram.

Phase Extraction from Optical Interferograms in Presence of Intensity Nonlinearity and Arbitrary Phase Shifts

One-shot Phase Retrieval Method for Interferometry Using a Hypercolumns Convolutional Neural Network

Phase Retrieval Using Iterative Fourier Transform and Convex Optimization Algorithm

Precise Phase Retrieval under Harsh Conditions by Constructing New Connected Interferograms

Phase retrieval for single-frame interferogram with an irregular-shaped aperture based on deep learning

A Simple Phase Retrieval Algorithm from a Single Shot Interferogram

Euclidean Matrix Norm Algorithm for Simultaneous Dual-Wavelength Phase-Shifting Interferometry with Blind Phase Shifts.