Abstract

Recently, deep learning (DL) has shown great potential in complex wavefront retrieval (CWR). However, the application of DL in CWR does not match well with the physical diffraction process. The state-of-the-art DL-based CWR methods crop full-size diffraction patterns down to a smaller size to save computational resources. However, cropping reduces the space–bandwidth product (SBP). In order to solve the trade-off between computational resources and SBP, we propose an imaging process matched neural network (IPMnet). IPMnet accepts full-size diffraction patterns with a larger SBP as inputs and retrieves a higher resolution and a larger field of view of the complex wavefront. We verify the effectiveness of the proposed IPMnet through simulations and experiments.

© 2023 Optica Publishing Group

Deep Gauss–Newton for phase retrieval

Kannara Mom, Max Langer, and Bruno Sixou
Opt. Lett. 48(5) 1136-1139 (2023)

Super-resolution multimode fiber imaging with an untrained neural network

Wei Li, Ksenia Abrashitova, and Lyubov V. Amitonova
Opt. Lett. 48(13) 3363-3366 (2023)

Improving the resolution of Fourier ptychographic imaging using an a priori neural network

Junting Sha, Wenmao Qiu, Guannan He, Zhi Luo, and Bo Huang
Opt. Lett. 48(23) 6316-6319 (2023)

Real-time phase imaging with physics-enhanced network and equivariance

Yuheng Wang, Huiyang Wang, Chengxin Zhou, Xianxin Han, Shengde Liu, Xiaoxu Lu, Jianglei Di, and Liyun Zhong
Opt. Lett. 48(10) 2732-2735 (2023)

High-frame-rate reconfigurable diffractive neural network based on superpixels

Yuan Qu, Hengyu Lian, Chunxu Ding, Huazhen Liu, Linxian Liu, and Jiamiao Yang
Opt. Lett. 48(19) 5025-5028 (2023)