Untrained deep network powered with explicit denoiser for phase recovery in inline holography

Abstract

Single-shot reconstruction of the inline hologram is highly desirable as a cost-effective and portable imaging modality in resource-constrained environments. However, the twin image artifacts, caused by the propagation of the conjugated wavefront with missing phase information, contaminate the reconstruction. Existing end-to-end deep learning-based methods require massive training data pairs with environmental and system stability, which is very difficult to achieve. Recently proposed deep image prior (DIP) integrates the physical model of hologram formation into deep neural networks without any prior training requirement. However, the process of fitting the model output to a single measured hologram results in the fitting of interference-related noise. To overcome this problem, we have implemented an untrained deep neural network powered with explicit regularization by denoising (RED), which removes twin images and noise in reconstruction. Our work demonstrates the use of alternating directions of multipliers method (ADMM) to combine DIP and RED into a robust single-shot phase recovery process. The use of ADMM, which is based on the variable splitting approach, made it possible to plug and play different denoisers without the need of explicit differentiation. Experimental results show that the sparsity-promoting denoisers give better results over DIP in terms of phase signal-to-noise ratio (SNR). Considering the computational complexities, we conclude that the total variation denoiser is more appropriate for hologram reconstruction.

Phase from defocus: transport of intensity or in-line holography?

“…Augmenting the number of images may enhance result accuracy but at the expense of increased measurement time. Notably, unlike TIE, DIHM has the potential for single-shot implementation 21 . Secondly, DIHM, being an interference technique, necessitates a coherent light source.…”

On the use of deep learning for phase recovery

“…: 1–180 (input only) l 2 -norm with aperture constraint and physical model Bai et al 154 Hologram dual-wavelength Phase CDD Expt. : 1 (input only) l 2 -norm with physical model Galande et al 155 Hologram Phase and amplitude U-Net Expt. : 1 (input only) l 2 -norm with physical model and denoiser Yao et al 159 3D diffraction image Phase and amplitude 3D Y-Net Sim.…”

“…Meanwhile, Bai et al 154 extended this from a single-wavelength case to a dual-wavelength case. Galande et al 155 found that this way of neural network optimization with a single-measurement intensity input lacks information diversity and can easily lead to overfitting of the noise, which can be mitigated by introducing an explicit denoiser. It is worth pointing out that this way of using the object-related intensity image as the neural network input makes it possible to internalize the mapping relationship between intensity and phase into the neural network through pre-training .…”

Generative adversarial neural network for 3D-hologram reconstruction

“…CNNs can be used for hologram generation [20][21][22] and reconstruction, including noise, twin image and zero-order suppression [23]. CNNs are typically used to reconstruct one image [24][25][26][27][28][29][30][31][32] or two (amplitude and phase information) [33][34][35][36][37][38][39][40][41] or extended focus imaging [41,42]. However, the direct reconstruction of the entire 3D-scene provides a wider range of possibilities [43].…”