【摘要】 由美国喷气推进实验室提出的MGS(modified G-S)算法,是一种基于迭代FFT变换的相位恢复波前传感方法。其中,多离焦面的使用保证算法迭代解具有唯一性,多种相位解包裹方法的引入使得算法既有很高的传感精度又有可观的传感动态范围。但当传感非连通域复杂光瞳时,由于受瞳内遮挡区域的伪包裹相位值影响,路径无关型解包裹算法失效,从而导致MGS算法迭代失败。为此首先对MGS算法做了进一步细节优化,在此基础上设计出一套普适于非连通域复杂光瞳的路径无关型解包裹算法流程,它能根据光瞳实际遮挡排布自适应设定最优解包裹步骤和路径。通过搭建实验光路,验证了该解包裹算法的可行性,并结合干涉测量对改进MGS算法在非连通域复杂光瞳中的波前传感精度做了初步标定。更多还原

【Abstract】 As a preferred wavefront sensing (WFS) method, the modified G-S (MGS) algorithm, which is classified as the iterative-transform algorithm of phase-diverse phase retrieval (PDPR), retrieves the phase distribution at pupil from some known point spread function (PSFs) at defocus planes. By using multiple diversities and proper phase unwrapping methods, this algorithm can produce unique phase estimation within a large dynamic range. But for nonconnected pupil, the path-independent phase unwrapping inside MGS is invalid owing to fake value of the points located in obstructed areas of pupil. Some detailed modifications about MGS are given. Then, a modified path-independent phase unwrapping algorithm is proposed based on the principle of least difference phase transition (LDPT), which can automatically optimize unwrapping steps and paths according to the pupil structure. Experiments have been implemented for MGS, and the adaptability of the proposed phase unwrapping algorithm has been validated. Compared with the measurement of interferometer, the WFS accuracy of MGS algorithm is calibrated.更多还原