Weighted constraint stochastic gradient descent algorithm for computational holographic near-eye display

Lizhi Chen; Runze Zhu; Songzhi Tian; Hao Zhang

doi:10.1117/12.2643564

19 December 2022 Weighted constraint stochastic gradient descent algorithm for computational holographic near-eye display

Lizhi Chen, Runze Zhu, Songzhi Tian, Hao Zhang

Author Affiliations +

Proceedings Volume 12318, Holography, Diffractive Optics, and Applications XII; 1231812 (2022) https://doi.org/10.1117/12.2643564
Event: SPIE/COS Photonics Asia, 2022, Online Only

Abstract

Stochastic gradient descent (SGD) algorithm with weighted constraint strategy is proposed to solve the vortex stagnation problem in CGH optimization and improve the image quality for computational holographic near-eye display. The weighted constraint strategy includes weighted phase constraint and weighted amplitude constraint. The weighted phase constraint is used to smooth the phase profile of reconstructed field, which helps to solve the vortex stagnation problem caused by optical vortices and eliminate the speckles in reconstructed field. The weighted amplitude constraint is used to broaden the optimization space by introducing the amplitude freedom of non-signal region in the reconstructed field, which helps to further improve the image quality in the signal region. The weighted constraint SGD algorithm can ensure the stable convergence of CGH optimization and avoid the vortex stagnation, which helps to eliminate the speckles and improve the image quality for holographic near-eye display.

Conference Presentation

Citation Download Citation

Lizhi Chen, Runze Zhu, Songzhi Tian, and Hao Zhang "Weighted constraint stochastic gradient descent algorithm for computational holographic near-eye display", Proc. SPIE 12318, Holography, Diffractive Optics, and Applications XII, 1231812 (19 December 2022); https://doi.org/10.1117/12.2643564

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