Laser & Optoelectronics Progress, Volume. 58, Issue 18, 1811020(2021)
Fourier Ptychography Based on Deep Learning
Fig. 1. Model of the FP forward imaging
Fig. 2. Reconstruction process of the FP image
Fig. 3. FP neural network based on physical model
Fig. 4. Simulation results of different algorithms. (a) INNM; (b) ePIE; (c) original high-resolution image; (d) sDR; (e) real image
Fig. 5. Simulation results of different aperture overlap ratios. (a) Real image; (b) 25%; (c) 50%; (d) 70%; (e) 85%
Fig. 6. Reconstruction results of ablation experiments. (a) Zernike polynomial and TV term are not introduced; (b) Zernike polynomial is introduced separately; (c) Zernike polynomial and TV term are introduced at the same time; (d) original high-resolution image
Fig. 7. Optical path of the macro FP
Fig. 8. Experimental setup of the macro FP
Fig. 9. Reconstruction result of USAF resolution target. (a) Low-resolution image; (b) enlarged detail 1 of the USAF; (c) ePIE; (d) INNM; (e) enlarged detail 2 of the USAF; (f) reference image
Get Citation
Copy Citation Text
hao Sha, Yangzhe Liu, Yongbing Zhang. Fourier Ptychography Based on Deep Learning[J]. Laser & Optoelectronics Progress, 2021, 58(18): 1811020
Category: Imaging Systems
Received: Jun. 2, 2021
Accepted: Jul. 20, 2021
Published Online: Aug. 28, 2021
The Author Email: Zhang Yongbing (ybzhang08@hit.edu.cn)