Related Experiment Video
Updated: Aug 18, 2025
07:26
Assessing Pupil-linked Changes in Locus Coeruleus-mediated Arousal Elicited by Trigeminal Stimulation
Published on: November 26, 2019
8.2K
Direct retrieval of Zernike-based pupil functions using integrated diffractive deep neural networks
Elena Goi1,2, Steffen Schoenhardt3,4, Min Gu5,6
1Institute of Photonic Chips, University of Shanghai for Science and Technology, Shanghai, 200093, China. elenagoi@usst.edu.cn.
Nature Communications
|December 8, 2022
Summary
We developed a new optical-electronic module using diffractive neural networks to directly retrieve pupil phase information. This compact system bypasses complex algorithms, enabling efficient phase retrieval for optical systems.
Area of Science:
- Optics and Photonics
- Computational Imaging
- Machine Learning Applications
Background:
- Pupil phase retrieval is crucial for optical systems like telescopes and microscopes.
- Existing methods require complex algorithms and computational resources.
- Accurate phase information is vital for aberration correction and imaging sensitive samples.
Purpose of the Study:
- To present a novel compact optical-electronic module for direct pupil phase retrieval.
- To demonstrate a method that bypasses traditional digital post-processing.
- To integrate diffractive neural networks with imaging sensors for real-time phase information extraction.
Main Methods:
- Utilized multi-layered diffractive neural networks printed on imaging sensors.
- Developed a system capable of directly retrieving Zernike-based pupil phase distributions.
- Validated the concept through numerical simulations and experimental demonstrations.
Main Results:
- Successfully demonstrated direct pupil phase retrieval of superpositions of the first 14 Zernike polynomials.
- Showcased the module's ability to extract phase information from an incident point spread function.
- Confirmed the effectiveness of the integrated optical-electronic approach.
Conclusions:
- The developed module offers a compact and efficient solution for pupil phase retrieval.
- Integration with CMOS sensors enables direct phase information extraction without post-processing.
- This technology holds potential for advancing optical system performance and imaging capabilities.
