Lensless phase retrieval based on deep learning used in holographic data storage

Abstract

This paper proposes a lensless phase retrieval method based on deep learning (DL) used in holographic data storage. By training an end-to-end convolutional neural network between the phase-encoded data pages and the corresponding near-field diffraction intensity images, the new unknown phase data page can be predicted directly from the intensity image by the network model without any iterations. The DL-based phase retrieval method has a higher storage density, lower bit-error-rate (BER), and higher data transfer rate compared to traditional iterative methods. The retrieval optical system is simple, stable, and robust to environment fluctuations which is suitable for holographic data storage. Besides, we studied and demonstrated that the DL method has a good suppression effect on the dynamic noise of the holographic data storage system.

Lensless complex amplitude demodulation based on deep learning in holographic data storage

“…Since the pioneering work of Sinha et al 24 on the recovery of phase directly from a single diffraction intensity image, deep learning networks have been used to reconstruct the amplitude and phase directly from a hologram 25−27 or combined with physics prior to retrieving the phase from a diffraction intensity image 28 . One can refer to a recent survey 29 for more details. In the existing studies, the retrieval of the complex amplitude from the hologram still needs reference beam.…”

Phase reconstruction based on deep learning with high pass filtering for holographic data storage

“…Phase-modulated holographic storage technology is favored because of its high coding rate and signal-to-noise ratio [3]. Since the phase information cannot be directly detected by the detector, how to accurately and quickly reconstruct the phase is the focus of phase-modulated holographic storage technology [4]. Generally, the phase reconstruction methods include interferometric method [5][6][7] and non-interferometric method [8][9].…”

“…Although the iterative phase reconstruction through dynamic sampling can reduce the number of iterations by about 3 times compared with the traditional iterative methods [12], the iterative method still hinders the improvement of transmission rate. Compared with the traditional iterative method, phase reconstruction technology based on deep learning is more suitable for holographic data storage because of its lower bit error rate and higher data transmission rate [4].…”

Bifocal-polarization holographic lens made from volume hologram

“…1. This unusual feature is widely used in polarization imaging [1][2][3][4] , microscopic amplification imaging [5,6] , stereo imaging [7][8][9][10] , projector [11] , etc., and may play an important role in the layered recording and reading of holographic discs [12][13][14] . In recent years, some bifocal polarization lens made of metasurface structures have been reported [15][16][17][18][19][20] .…”