Implementation of Convolutional Neural Networks in a Virtual Optical System for Wavefront Detection with Potential Application in Visual Optics

Authors

  • Andrés Osorno-Quiroz Programa de Ingeniería de software y datos, Facultad de Ingeniería y Ciencias Agropecuarias, Institución Universitaria Digital de Antioquia, Medellín, Colombia
  • Walter Torres-Sepúlveda Programa de Ciencias Ambientales, Grupo de Investigación en Innovación Digital y Desarrollo Social (INDDES), Facultad de Ciencias y Humanidades, Institución Universitaria Digital de Antioquia, Medellín, Colombia

DOI:

https://doi.org/10.4302/plp.v16i3.1285

Abstract

In this work, the effectiveness of the convolutional neural network architectures AlexNet and a new proposed ResNet-based architecture are compared for the detection of optical aberrations in typical images from a Hartmann-Shack sensor, within the range of values associated with the average aberrations of a real eye. Both neural networks are trained with a dataset built from a virtual optical system containing more than 44,000 training, validation, and testing images. The results demonstrated that both neural networks were able to accurately predict the typical aberrations simulated for a real eye, with better performance for the proposed ResNet CNN. Compared to traditional methods such as the centroid detection method in Hartmann-Shack images, this artificial intelligence-based approach presents itself as an effective and promising alternative for aberration detection where there are no such restrictive conditions as dynamic range, making this methodology and the proposed ResNet potentially applicable in fields such as adaptive optics and ophthalmology.

Full Text: PDF

References

  1. J. C. Ochoa-Tabares, E. Hernández-Quíntela, N. D. C. Ruiz-Quintero, R. Naranjo-Tackman. "Precisión de un sistema de aberrometría hartmann-shack, en la evaluación de las aberraciones oculares", Revista Mexicana de Oftalmologia. 86, 25 (2012). DirectLink
  2. J. Li, Y. Gong, X. Hu, and C. Li., "A High-Precision Centroid Detecting Method for Hartmann-Shack Wavefront Sensor", Chin. J. Lasers. 41(3), 316002 (2014). CrossRef
  3. H. Guo, N. Korablinova, Q. Ren, and J. Bille. "Wavefront reconstruction with artificial neural networks", Opt. Express 14, 6456 (2006). CrossRef
  4. S. Barwick. "Detecting higher-order wavefront errors with an astigmatic hybrid wavefront sensor", Opt. Lett. 34, 1690 (2009). CrossRef
  5. Z. Li and X. Li. "Centroid computation for Shack-Hartmann wavefront sensor in extreme situations based on artificial neural networks", Opt. Express 26, 31675 (2018). CrossRef
  6. T. B. DuBose, D. F. Gardner, and A. T. Watnik. "Intensity-enhanced deep network wavefront reconstruction in Shack–Hartmann sensors", Opt. Lett. 45, 1699 (2020). CrossRef
  7. S. Plainis, and I. G. Pallikaris, "Ocular monochromatic aberration statistics in a large emmetropic population", J. Mod. Opt. 55, 759 (2008). CrossRef
  8. H. Zhang, J. Zhao, H. Chen et al. "Large-Dynamic-Range Ocular Aberration Measurement Based on Deep Learning with a Shack–Hartmann Wavefront Sensor", Sensors. 24, 2728 (2024). CrossRef
  9. B. Zhang, J. Zhu, K. Si, W. Gong. "Deep Learning Assisted Zonal Adaptive Aberration Correction", Front. Phys. 8, 621966 (2021). CrossRef
  10. J. Jiang, X. Liu, Y. Chen, S. Gao, Y. Liu, Y. Jiang. "Aberrations estimate based on diffraction patterns using deep residual network", AIP Adv. 13, 105322 (2023). CrossRef

Downloads

Published

2024-10-01

How to Cite

[1]
A. Osorno-Quiroz and W. Torres-Sepúlveda, “Implementation of Convolutional Neural Networks in a Virtual Optical System for Wavefront Detection with Potential Application in Visual Optics”, Photonics Lett. Pol., vol. 16, no. 3, pp. 49–51, Oct. 2024.

Issue

Vol. 16 No. 3 (2024)

Section

Articles

License

Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).