Xuyu Zhang, Shengfu Cheng, Jingjing Gao, Yu Gan, Chunyuan Song, Dawei Zhang, Songlin Zhuang, Shensheng Han, Puxiang Lai, Honglin Liu, "Physical origin and boundary of scalable imaging through scattering media: a deep learning-based exploration," Photonics Res. 11, 1038 (2023)

Search by keywords or author
- Photonics Research
- Vol. 11, Issue 6, 1038 (2023)

Fig. 1. Schematic of the experimental setup of imaging through a diffuser with the coordinate system labeled. The insets (a) and (b) show the settings of imaging regions to acquire the training and test data in Tests I and II, respectively. In Test I, the training data were obtained from region A (red circle) only with the test data from regions 1–5 (green circles). In Test II, the training data were obtained from regions A – E (red circles) with the test data from regions 1–5 (green circles).

Fig. 2. Experimental results. (a) Image reconstruction through a homemade diffuser in Tests I and II. (b) Curves of the averaged PCC with error bar for 10 reconstructed images at each of the test regions 1–5. Note a nonuniform abscissa is adopted to better reflect the whole trend, given the nonuniform distributed displacements. (c) The CCC curve measured in experiment with a FWHM of ∼ 34 μm .

Fig. 3. Results of Simulation I where no ballistic light is involved. (a) The phase map of the simulated diffuser in which the color bar denotes the range of phase value in radian. (b) The characterized CCC curve of the simulated diffuser, which has an FWHM of ∼ 36 μm and a base level of zero. (c) The speckle patterns and predicted images in both Tests I and II at regions 1, 3, 5, 7, and 9 with the ground truth on the left. (d) Curves of averaged PCC with error bar for two tests in Simulation I in which experimental PCC results are also included for comparison. The right subplot shows the zoom-in area of the dash rectangle.

Fig. 4. Results of Simulation II that involves different weights of ballistic light (η ). (a) Phase distributions of the simulated diffusers corresponding to different value of η . (b) The image reconstruction results on test regions 1–6 of varying Δ x when the network is trained under different η . Note that rows I–VI correspond to η = 0.1 , 0.3 , 0.5 , 0.7 , 0.9 , a n d 1 , respectively. (c) The curves of average PCC as a function of displacement for different η . (d) The CCC curves of output field under different η .

Fig. 5. Improved model generalization by increasing distance z 3 in experiment. (a) The CCC curves measured experimentally for z 3 = 5 , 10 , 20 cm , respectively. (b) The curves of average PCC for network testing on a series of regions in experimental Tests I and II under the case of different z 3 .

Set citation alerts for the article
Please enter your email address