Vol. 53, Issue 1, pp. 127-140

Vol. 53 Issue 1 pp. 127-140

Phase retrieval without phase unwrapping for white blood cells in deep-learning phase-shifting digital holography

Shuyang Jin, Xiaoqing Xu, Jili Chen, Yudan Ni


digital holography, phase retrieval, deep learning


Phase retrieval and phase unwrapping are the two important problems for enabling quantitative phase imaging of cells in phase-shifting digital holography. To simultaneously cope with these two problems, a deep-learning phase-shifting digital holography method is proposed in this paper. The proposed method can establish the continuous mapping function of the interferogram to the ground-truth phase using the end-to-end convolutional neural network. With a well-trained deep convolutional neural network, this method can retrieve the phase from one-frame blindly phase-shifted interferogram, without phase unwrapping. The feasibility and applicability of the proposed method are verified by the simulation experiments of the microsphere and white blood cells, respectively. This method will pave the way to the quantitative phase imaging of biological cells with complex substructures.

Vol. 53
Issue 1
pp. 127-140

1.28 MB
OPTICA APPLICATA - a quarterly of the Wrocław University of Science and Technology, Faculty of Fundamental Problems of Technology