Artificial intelligence and Fourier optics: Application of DeepLabV3+ in the recovery of a diffracting aperture in light propagation


  • Cesar Camacho-Bello Universidad Politécnica de Tulancingo
  • Lucia Gutierrez-Lazcano Universidad Politécnica de Tulancingo
  • Rosa Ortega-Mendoza Universidad Politécnica de Tulancingo



Diffraction, Optical Systems, Deep Learning


The combination of Fourier Optics and Artificial Intelligence has driven significant advances in image processing and modeling of optical systems, with the UNet architecture being the main protagonist. However, the DeepLabV3+ network has recently shown promising performance in detecting transfer opens. In this study, we investigate the effectiveness of DeepLabV3+ in identifying transfer apertures in light propagation models and compare its performance with that of UNet. The results reveal that DeepLabV3+ outperforms UNet in terms of accuracy and robustness in identifying transfer apertures, even in the presence of noise and aperture shape variations.


E. Hecht, Optics, 5th ed. (Pearson, 2017)

J. W. Goodman, Fourier Optics, 3rd ed. (Roberts and Company Publishers, 2005)

L. Wang and H. Wu, Biomedical Optics: Principles and Imaging, 10-15 John Wiley & Sons, New Jersey (2007)

D. Yue, Y. He, and Y. Li, Piston error measurement for segmented telescopes with an artificial neural network, Sensors 21 (2021) 3364

U. Flechsig et al., Physical optics simulations with PHASE for SwissFEL beamlines, In AIP Conference Proceedings, vol. 1741 (AIP Publishing, 2016)

N. Siddique et al., U-Net and Its Variants for Medical Image Segmentation: A Review of Theory and Applications, IEEE Access 9 (2021) 82031,

T. Zeng, Y. Zhu, and E. Y. Lam, Deep learning for digital holography: a review, Opt. Express 29 (2021) 40572,

Z. Zhang et al., Holo-UNet: hologram-to-hologram neural network restoration for high fidelity low light quantitative phase imaging of live cells, Biomed. Opt. Express 11 (2020) 5478,

J. Wu et al., High-speed computer-generated holography using an autoencoder-based deep neural network, Opt. Lett. 46 (2021) 2908,

T. Zhang et al., Rapid and robust two-dimensional phase unwrapping via deep learning, Opt. Express 27 (2019) 23173,

C.Wu et al., Phase unwrapping based on a residual en-decoder network for phase images in Fourier domain Doppler optical coherence tomography, Biomed. Opt. Express 11 (2020) 1760,

X. Dun et al., Learned rotationally symmetric diffractive achromat for full-spectrum computational imaging, Optica 7 (2020) 913,

Z. Wei et al., Elimination of stripe artifacts in light sheet fluorescence microscopy using an attention-based residual neural network, Biomed. Opt. Express 13 (2022) 1292,

O. Ronneberger et al., Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015 (Springer International Publishing, Cham, 2015) pp. 234-241, 28

L.-C. Chen et al., Rethinking atrous convolution for semantic image segmentation, arXiv preprint arXiv:1706.05587 (2017)

L.-C. Chen et al., Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation, In Computer Vision - ECCV 2018: 15th European Conference, Munich, Germany, September 8-14, 2018, Proceedings, Part VII (SpringerVerlag, Berlin, Heidelberg, 2018) p. 833,

K. He and J. Sun, Convolutional neural networks at constrained time cost, In 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2015) pp. 5353,




How to Cite

C. Camacho-Bello, L. Gutierrez-Lazcano, and R. Ortega-Mendoza, “Artificial intelligence and Fourier optics: Application of DeepLabV3+ in the recovery of a diffracting aperture in light propagation”, Rev. Mex. Fís., vol. 70, no. 1 Jan-Feb, pp. 011301 1–, Jan. 2024.