Physics informed neural network with Fourier basis for computer-generated hologram synthesizing

Runze Zhu; Lizhi Chen; Hao Zhang

doi:10.1117/12.2686158

27 November 2023 Physics informed neural network with Fourier basis for computer-generated hologram synthesizing

Runze Zhu, Lizhi Chen, Hao Zhang

Proceedings Volume 12768, Holography, Diffractive Optics, and Applications XIII; 127680Y (2023) https://doi.org/10.1117/12.2686158
Event: SPIE/COS Photonics Asia, 2023, Beijing, China

Abstract

Deep learning is an emerging technique for computer hologram generation, and the training dataset is critical for the quality of the learning-based hologram. In this study, we propose a phase hologram generation method based on a deep neural network (DNN) trained with Fourier basis functions. As the training dataset is generated by recombining the Fourier basis functions, its statistical frequency characteristics can be directly manipulated to improve the performance of DNN. With this physics informed training strategy, the external generalization of the model, as well as the image quality of the holographic reconstructions, can be improved. Experiments demonstrate that the proposed method is effective for achieving high-quality optical reconstructions.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Runze Zhu, Lizhi Chen, and Hao Zhang "Physics informed neural network with Fourier basis for computer-generated hologram synthesizing", Proc. SPIE 12768, Holography, Diffractive Optics, and Applications XIII, 127680Y (27 November 2023); https://doi.org/10.1117/12.2686158

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available