基于多波长同轴数字全息的相位恢复算法

同轴数字全息中固有的共轭像问题会严重影响再现像质量,已有的相位恢复算法需要大量计算才能祛除共轭像,且无法获得连续分布的解包裹相位像。提出一种基于多波长的相位恢复算法,该算法利用4个波长下记录的数字全息图,并在迭代过程中借助光学解包裹原理扩展了测量范围,实现了再现成像面中共轭像祛除的同时获得解包裹的相位分布。同时,由于在物面及记录面添加了约束条件,该方法中所用波长数少于其他已提出的基于多波长原理的相位恢复算法,而且具有更快的迭代收敛速度和更好的共轭像祛除效果。数值仿真和实验结果皆证明了该方法的有效性。

Phase retrieval for multiple-wavelength in-line digital holography

In-line digital holography, because of its most simplified holographic recording architecture without extra reference beam, has a larger field of view compared with off-axis digital holography. Thus, it has been applied in many fields. However, both the defocused conjugate images and zero-order autocorrelation are superimposed on the real image of the object in the reconstruction process, which blurs the real image and degrades the measurement accuracy. A phase retrieval method is proposed to eliminate the unwanted conjugate image based on four diffraction intensity patterns recorded at different illumination wavelengths for in-line digital holography, in which the appropriate constraints are imposed both on the hologram plane and object plane. A synthetic wavelength is used to retrieve wrap-free phase distributions with a much extended vertical measurement range, thus the reconstructed phase distribution directly provides the height distribution of the surface of the sample after eliminating the twin image. In comparison with previous research, the proposed method requires a fewer number of illumination wavelengths and has a faster rate of convergence, as well as a better elimination effect, which is demonstrated by the simulation and experimental results.