Fractional Vortex Beam Modes Recognition Based on I-ResNet Network

Dongmei Wei; Qian Du; Fangning Liu; Ke Wang; Yuefeng Zhao

doi:10.3788/AOS231361

Acta Optica Sinica, Volume. 43, Issue 23, 2326001(2023)

Fractional Vortex Beam Modes Recognition Based on I-ResNet Network

Dongmei Wei, Qian Du, Fangning Liu, Ke Wang, and Yuefeng Zhao^*

Shandong Provincial Engineering and Technical Center of Light Manipulations, School of Physics and Electronics, Shandong Normal University, Jinan 250358, Shandong , China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(14)

Fig. 1. Schematic diagram of beam transmission through phase screen

Download full size

Fig. 2. Simulation of atmospheric turbulence phase screen at different intensities.(a) $C_{n}^{2} = 10^{- 16} m^{- 2 / 3}$ ; (b) $C_{n}^{2} = 10^{- 15} m^{- 2 / 3}$ ; (c) $C_{n}^{2} = 10^{- 14} m^{- 2 / 3}$

Download full size

$Influence of atmospheric turbulence intensity on intensity distribution of fractional vortex beam$

Fig. 3. Influence of atmospheric turbulence intensity on intensity distribution of fractional vortex beam

Download full size

Fig. 4. Structure diagram of I-ResNet network

Download full size

Fig. 5. Residual structure. (a) ResNet50 residual block; (b) I-ResNet residual block

Download full size

Fig. 6. Influence of mode resolution on mode detection. (a) Δz=500 m; (b) Δz=1000 m; (c) Δz=1500 m

Download full size

Fig. 7. Influence of mode integer part on mode detection

Download full size

Fig. 8. Accuracy rate and loss rate varying with number of iterations. (a) Accuracy rate varying with number of iterations; (b) loss rate varying with number of iterations

Download full size

Fig. 9. Testing set accuracy

Download full size

Fig. 10. Testing set confusion matrix

Download full size

Table 1. Five mode resolutions and corresponding mode ranges
View table
Table 1. Five mode resolutions and corresponding mode ranges
（Mode resolution） $Δ l$ 0.05 0.10 0.15 0.20 0.25
Mode range 5.10-5.55 5.10-6.00 5.10-6.45 5.10-6.90 5.10-7.35

Table 2. Beam dataset of $Δ l = 0.1$
View table
Table 2. Beam dataset of $Δ l = 0.1$
$Δ l = 0.1$ $C_{n}^{2} / m^{- 2 / 3}$ △z /m Sample number Total number
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
6.0
$1 \times 10^{- 16}$ 500 300 3000
1000 300 3000
1500 300 3000
$1 \times 10^{- 15}$ 500 300 3000
1000 300 3000
1500 300 3000
$1 \times 10^{- 14}$ 500 300 3000
1000 300 3000
1500 300 3000

Table 3. Integer parts of l and corresponding mode ranges
View table
Table 3. Integer parts of l and corresponding mode ranges
（Integer part of l） $l_{i n t e g e r}$ 1 2 5
Mode range 1.10-2.00 2.10-3.00 5.10-6.00

Table 4. Comparison of mode recognition results of ResNet50 and I-ResNet networks
View table
Table 4. Comparison of mode recognition results of ResNet50 and I-ResNet networks
$C_{n}^{2}$ $1 \times 10^{- 16} m^{- 2 / 3}$ $1 \times 10^{- 15} m^{- 2 / 3}$ $1 \times 10^{- 14} m^{- 2 / 3}$
ResNet50 100% 99.8% 83.4%
I-ResNet 100% 100% 89.5%

Tools

Get Citation

Copy Citation Text

Dongmei Wei, Qian Du, Fangning Liu, Ke Wang, Yuefeng Zhao. Fractional Vortex Beam Modes Recognition Based on I-ResNet Network[J]. Acta Optica Sinica, 2023, 43(23): 2326001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites