Hierarchical phase unwrapping for dual-wavelength digital holographic microscopy

Abstract

Dual-wavelength digital holographic microscopy (DHM) can extend ambiguity range in phase unwrapping by synthesizing a much longer beat wavelength, but it suffers from phase error amplification that destroys the vertical sensitivity and accuracy. To achieve a better vertical sensitivity and avoid height jump phenomenon, a hierarchical method is presented to construct a wavelength chain and unwrap the phase gradually from long to short wavelengths. This method could improve the vertical sensitivity by synthesizing a shorter wavelength and avoid height jump by reducing the estimation error of integer part of phase order. Theoretical analysis, simulations, and experiments demonstrate the effectiveness of the hierarchical method to attenuate the height jump and preserve vertical sensitivity in dual-wavelength DHM phase unwrapping. The method is simple and deterministic, and only costs a few milliseconds for unwrapping a typical phase map of mega pixels.

Depth_ASPP_UNet++: a novel 2D ESPI phase unwrapping based on convolutional neural network

“…However, no matter which phase shift technique is used, the collected phase information is obtained by calculating the inverse trigonometric function, resulting in phase values wrapped within the range of [−π,π]. The wrapped phase is displayed as a step discontinuous phase distribution pattern [12,13]. To obtain the true surface topography of the measured object, the wrapped phase must undergo the phase unwrapping process [14,15].…”