Single Measurement Surface 2-D Near-Field Phase Recovery via Smart Warping Sampling and Factorization of 1-D Polynomials

We propose a new approach to phase retrieval (PR) from near-field measurements. By jointly exploiting a smart warping sampling strategy as well as recent advances in the far-field case, the technique enables an optimal solution of the problem by exploiting only one measurement surface and a single probe. This overcomes a common limitation of phaseless near-field techniques, which typically require two distinct datasets. Additionally, the proposed deterministic approach avoids resource-intensive global optimization algorithms, prevents false solutions, and minimizes the number of required measurements. The given theory is supported by several examples, including experimental data and an analysis of the method’s sensitivity to noise.