Phase unwrapping is an important but challenging issue in phase measurement. Even with the research efforts of a few decades, unfortunately, the problem remains not well solved, especially when heavy noise and aliasing (undersampling) are present. We propose a database generation method for phase-type objects and a one-step deep learning phase unwrapping method. With a trained deep neural network, the unseen phase fields of living mouse osteoblasts and dynamic candle flame are successfully unwrapped, demonstrating that the complicated nonlinear phase unwrapping task can be directly fulfilled in one step by a single deep neural network. Excellent anti-noise and anti-aliasing performances outperforming classical methods are highlighted in this paper.

In this Letter, for the first time, to the best of our knowledge, we propose a digital holographic reconstruction method with a one-to-two deep learning framework (Y-Net). Perfectly fitting the holographic reconstruction process, the Y-Net can simultaneously reconstruct intensity and phase information from a single digital hologram. As a result, this compact network with reduced parameters brings higher performance than typical network variants. The experimental results of the mouse phagocytes demonstrate the advantages of the proposed Y-Net.

Estimation of absolute (true) phase is an important but challenging problem in many imaging or measurement techniques, such as optical interferometry (OI),¹ magnetic resonance imaging (MRI),² fringe projection profilometry (FPP),^3,4 and interferometric synthetic aperture radar (InSAR)^5,6 (see Fig. 1).

Deep learning neural networks are used for wavefront sensing and aberration correction in atmospheric turbulence without any wavefront sensor (i.e. reconstruction of the wavefront aberration phase from the distorted image of the object). We compared and found the characteristics of the direct and indirect reconstruction ways: (i) directly reconstructing the aberration phase; (ii) reconstructing the Zernike coefficients and then calculating the aberration phase. We verified the generalization ability and performance of the network for a single object and multiple objects. What’s more, we verified the correction effect for a turbulence pool and the feasibility for a real atmospheric turbulence environment.

Theoretical analysis shows that, to improve the resolution and the range of the field of view of the reconstructed image in digital lensless Fourier transform holography, an effective solution is to increase the area and the pixel number of the recorded digital hologram. A new approach based on the synthetic aperture technique and use of linear CCD scanning is presented to obtain digital holographic images with high resolution and a wide field of view. By using a synthetic aperture technique and linear CCD scanning, we obtained digital lensless Fourier transform holograms with a large area of $3.5\mathrm{cm}\times 3.5\mathrm{cm}$ ($5000\times 5000$ pixels). The numerical reconstruction of a $4\mathrm{mm}$ object at a distance of $14\mathrm{cm}$ by use of a Rayleigh–Sommerfeld integral shows that a theoretically minimum resolvable distance of $2.57\mathrm{\mu m}$ can be achieved at a wavelength of $632.8\mathrm{nm}$. The experimental results are consistent with the theoretical analysis.