x

## 留言板

Extracting atmospheric turbulence phase using deep convolutional neural network

## Extracting atmospheric turbulence phase using deep convolutional neural network

Xu Qi-Wei, Wang Pei-Pei, Zeng Zhen-Jia, Huang Ze-Bin, Zhou Xin-Xing, Liu Jun-Min, Li Ying, Chen Shu-Qing, Fan Dian-Yuan
PDF
HTML
• #### Abstract

When a light beam transmits in free space, it is easily affected by atmospheric turbulence. The effect on transmitted light is equivalent to adding a random noise phase to it, which leads its transmission quality to deteriorate. The method of improving the quality of transmitted beams is usually to compensate for the phase distortion at the receiver by adding reverse turbulence phase, and the premise of this method is to obtain the turbulence phase carried by the distorted beam. The adaptive optics system is the most common way to extract the phase information. However, it is inefficient to be applied to varying turbulence environments due to the fact that a wave-front sensor and complex optical system are usually contained. Deep convolutional neural network (CNN) that can directly capture feature information from images is widely used in computer vision, language processing, optical information processing, etc. Therefore, in this paper proposed is a turbulence phase information extraction scheme based on the CNN, which can quickly and accurately extract the turbulence phase from the intensity patterns affected by atmosphere turbulence. The CNN model in this paper consists of 17 layers, including convolutional layers, pooling layers and deconvolutional layers. The convolutional layers and pooling layers are used to extract the turbulent phase from the feature image, which is the core structure of the network. The function of the deconvolutional layers is to visualize the extracted turbulence information and output the final predicted turbulence phase. After learning a huge number of samples, the loss function value of CNN converges to about 0.02, and the average loss function value on the test set is lower than 0.03. The trained CNN model has a good generalization capability and can directly extract the turbulent phase according to the input light intensity pattern. Using an I5-8500 CPU, the average time to predict the turbulent phase is as low as s under the condition of $C_{{n}}^2 = 1 \times {10^{ - 14}}\;{{\rm{m}}^{ - 2/3}}$, $5 \times {10^{ - 14}}\;{{\rm{m}}^{ - 2/3}}$, and $1 \times {10^{ - 13}}\;{{\rm{m}}^{ - 2/3}}$. In addition, the turbulence phase extraction capability of CNN can be further enhanced by improving computing power or optimizing model structure. These results indicate that the CNN-based turbulence phase extraction method can effectively extract the turbulence phase, which has important application value in turbulence compensation, atmospheric turbulence characteristics research and image reconstruction.

#### References

 [1] Zheng G, Wang L, Wang J, Zhou M C, Song M M 2018 J. Mod. Opt. 65 1616 [2] Yuan Y S, Liu D, Zhou Z X, Xu H F, Qu J, Cai Y J 2018 Opt. Express 26 21861 [3] Li Y Q, Wang L G, Wu Z S 2018 Optik 158 1349 [4] Wang Y K, Xu H Y, Li D Y, Wang R, Jin C B, Yin X H, Gao S J, Mu Q Q, Xuan L, Cao Z L 2018 Sci. Rep. 8 1124 [5] Gerçekcioğlu H 2019 Opt. Commun. 439 233 [6] Usenko V C, Peuntinger C, Heim B, Günthner K, Derkach I, Elser D, Marquardt C, Filip R, Leuchs G 2018 Opt. Express 26 31106 [7] Hope D A, Jefferies S M, Hart M, Nagy J G 2016 Opt. Express 24 12116 [8] Wen W, Jin Y, Hu M J, Liu X L, Cai Y J, Zou C J, Luo M, Zhou L W, Chu X X 2018 Opt. Commun. 415 48 [9] Ren Y X, Xie G D, Huang H, Ahmed N, Yan Y, Li L, Bao C J, Lavery M P, Tur M, Neifeld M A, Boyd R W, Shapiro J H, Willner A E 2014 Optica 1 376 [10] Yin X L, Chang H, Cui X Z, Ma J X, Wang Y J, Wu G H, Zhang L J, Xin X J 2018 Appl. Opt. 57 7644 [11] Neo R, Goodwin M, Zheng J, Lawrence J, Leon-Saval S, Bland-Hawthorn J, Molina-Terriza G 2016 Opt. Express 24 2919 [12] Gerchberg R W 1972 Optik 35 237 [13] Fu S Y, Zhang S K, Wang T L, Gao C Q 2016 Opt. Lett. 41 3185 [14] Nelson W, Palastro J P, Wu C, Davis C C 2016 Opt. Lett. 41 1301 [15] Hinton G E, Salakhutdinov R R 2006 Science 313 504 [16] Lecun Y, Bengio Y, Hinton G 2015 Nature 521 436 [17] Li J, Zhang M, Wang D S, Wu S J, Zhan Y Y 2018 Opt. Express 26 10494 [18] Roddier N A 1990 Opt. Eng. 29 1174 [19] Mcglamery B L 1967 J. Opt. Soc. Am. 57 293 [20] Zhao S M, Leach J, Gong L Y, Ding J, Zheng B Y 2012 Opt. Express 20 452 [21] Rumerlhar D E 1986 Nature 323 533 [22] Lecun Y, Bottou L, Bengio Y, Haffner P 1998 Proc. IEEE 86 2278 [23] Barakat R, Newsam G 1985 J. Opt. Soc. Am. A 2 2027 [24] Guo Y M, Liu Y, Oerlemans A, Lao S Y, Wu S, Lew M S 2016 Neurocomputing 187 27 [25] Hinton G E, Osindero S, Teh Y 2006 Neural Comput. 18 1527 [26] Qian Y M, Bi M X, Tan T, Yu K 2016 IEEE Trans. Audio Speech Lang. Process. 24 2263 [27] Sheridan P M, Cai F X, Du C, Ma W, Zhang Z Y, Lu W D 2017 Nat. Nanotech. 12 784

#### Cited By

• 图 1  各湍流强度下的随机相位屏　(a), (b) $C_{{n}}^2 \!=\! 1 \!\!\times\!\! {10^{ - 14}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$; (c), (d) $C_{{n}}^2 \!=\! 5 \!\!\times\!\!{10^{ - 14}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$; (e), (f) $C_{\rm{n}}^2 \!=\! 1 \!\!\times\!\! {10^{ - 13}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$

Figure 1.  Random phase screen at each turbulence intensity: (a), (b) $C_{{n}}^2 = 1 \times {10^{ - 14}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$; (c), (d) $C_{\rm{n}}^2 = 5 \times {10^{ - 14}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$; (e), (f) $C_{\rm{n}}^2 = 1 \times {10^{ - 13}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$.

图 2  各湍流强度影响下传输光束截面光斑　(a)初始高斯光束; (b), (c) $C_{\rm{n}}^2 = 1 \times {10^{ - 14}}\;{{\rm{m}}^{ - 2/3}}$; (d), (e) $C_{\rm{n}}^2 = 5 \times$ ${10^{ - 14}}\;{{\rm{m}}^{ - 2/3}}$; (f), (g) $C_{\rm{n}}^2 = 1 \times {10^{ - 13}}\;{{\rm{m}}^{ - 2/3}}$

Figure 2.  The cross-section spot of transmission beam at each turbulence intensity: (a) Initial Gaussian beam; (b), (c) $C_{\rm{n}}^2 =$ $1 \times {10^{ - 14}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$; (d), (e) $C_{\rm{n}}^2 = 5 \times {10^{ - 14}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$; (f), (g) $C_{\rm{n}}^2 = 1 \times {10^{ - 13}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$.

图 3  提取湍流相位的CNN结构

Figure 3.  The CNN structure of extracting the turbulent phase.

图 4  训练过程损失函数曲线

Figure 4.  The loss function curve of training process.

图 5  训练过程提取到的湍流相　(a)与(b), (c)与(d), (e)与(f), (g)与(h), (i)与(j)和(k)与(l)的迭代次数分别为1, 100, 500, 4000, 8000, 14000

Figure 5.  The turbulent phase during the training process: The number of iterations of (a) and (b), (c) and (d), (e) and (f), (g) and (h), (i) and (j), and (k) and (l) is 1, 100, 500, 4000, 8000, 14000.

图 6  各湍流强度损失函数曲线

Figure 6.  The loss function curve at each turbulence intensity.

图 7  不同湍流强度下, 经过CNN提取到的湍流相位　(a), (b), (c)初始高斯光束; (d), (e), (f) 受大气湍流影响的高斯光束; (g), (h), (i)实际的大气湍流相位; (j), (k), (l) CNN输出的预测湍流相位

Figure 7.  The predicted turbulent phase based on CNN at each turbulence intensity: (a), (b), (c) Initial Gaussian beam; (d), (e), (f) Gaussian beam affected by atmospheric turbulence; (g), (h), (i) the actual turbulence phase; (j), (k), (l) the output phase of CNN.

图 8  CNN与GS算法提取湍流相位效果对比　(a), (b), (c)受湍流强度为$C_{\rm{n}}^2 = 1 \times {10^{ - 14}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$影响的高斯光束; (d), (e), (f)实际湍流相位; (g), (h), (i)基于CNN模型提取的湍流相位; (j), (k), (l) GS算法提取的湍流相位

Figure 8.  The comparison of CNN and GS algorithm for extracting turbulence phase: (a), (b), (c) Gaussian beam affected by atmospheric turbulence with $C_{\rm{n}}^2 = 1 \times {10^{ - 14}}\;{{\rm{m}} ^{{{ - 2} / 3}}}$; (d), (e), (f) the actual turbulence phase; (g), (h), (i) the predicted turbulent phase based on CNN; (j), (k), (l) the predicted turbulent phase based on GS algorithm.

图 9  训练及验证测试结果　(a)训练过程损失函数曲线; (b)训练过程中利用验证集测试得到的预测湍流相位

Figure 9.  Training and validation set test results: (a) The loss function curve of training process; (b) the predicted turbulence phase obtained by testing the validation set during training.

•  [1] Zheng G, Wang L, Wang J, Zhou M C, Song M M 2018 J. Mod. Opt. 65 1616 [2] Yuan Y S, Liu D, Zhou Z X, Xu H F, Qu J, Cai Y J 2018 Opt. Express 26 21861 [3] Li Y Q, Wang L G, Wu Z S 2018 Optik 158 1349 [4] Wang Y K, Xu H Y, Li D Y, Wang R, Jin C B, Yin X H, Gao S J, Mu Q Q, Xuan L, Cao Z L 2018 Sci. Rep. 8 1124 [5] Gerçekcioğlu H 2019 Opt. Commun. 439 233 [6] Usenko V C, Peuntinger C, Heim B, Günthner K, Derkach I, Elser D, Marquardt C, Filip R, Leuchs G 2018 Opt. Express 26 31106 [7] Hope D A, Jefferies S M, Hart M, Nagy J G 2016 Opt. Express 24 12116 [8] Wen W, Jin Y, Hu M J, Liu X L, Cai Y J, Zou C J, Luo M, Zhou L W, Chu X X 2018 Opt. Commun. 415 48 [9] Ren Y X, Xie G D, Huang H, Ahmed N, Yan Y, Li L, Bao C J, Lavery M P, Tur M, Neifeld M A, Boyd R W, Shapiro J H, Willner A E 2014 Optica 1 376 [10] Yin X L, Chang H, Cui X Z, Ma J X, Wang Y J, Wu G H, Zhang L J, Xin X J 2018 Appl. Opt. 57 7644 [11] Neo R, Goodwin M, Zheng J, Lawrence J, Leon-Saval S, Bland-Hawthorn J, Molina-Terriza G 2016 Opt. Express 24 2919 [12] Gerchberg R W 1972 Optik 35 237 [13] Fu S Y, Zhang S K, Wang T L, Gao C Q 2016 Opt. Lett. 41 3185 [14] Nelson W, Palastro J P, Wu C, Davis C C 2016 Opt. Lett. 41 1301 [15] Hinton G E, Salakhutdinov R R 2006 Science 313 504 [16] Lecun Y, Bengio Y, Hinton G 2015 Nature 521 436 [17] Li J, Zhang M, Wang D S, Wu S J, Zhan Y Y 2018 Opt. Express 26 10494 [18] Roddier N A 1990 Opt. Eng. 29 1174 [19] Mcglamery B L 1967 J. Opt. Soc. Am. 57 293 [20] Zhao S M, Leach J, Gong L Y, Ding J, Zheng B Y 2012 Opt. Express 20 452 [21] Rumerlhar D E 1986 Nature 323 533 [22] Lecun Y, Bottou L, Bengio Y, Haffner P 1998 Proc. IEEE 86 2278 [23] Barakat R, Newsam G 1985 J. Opt. Soc. Am. A 2 2027 [24] Guo Y M, Liu Y, Oerlemans A, Lao S Y, Wu S, Lew M S 2016 Neurocomputing 187 27 [25] Hinton G E, Osindero S, Teh Y 2006 Neural Comput. 18 1527 [26] Qian Y M, Bi M X, Tan T, Yu K 2016 IEEE Trans. Audio Speech Lang. Process. 24 2263 [27] Sheridan P M, Cai F X, Du C, Ma W, Zhang Z Y, Lu W D 2017 Nat. Nanotech. 12 784
•  [1] Fan Shuang, Zhang Ya-Ping, Wang Fan, Gao Yun-Long, Qian Xiao-Fan, Zhang Yong-An, Xu Wei, Cao Liang-Cai. Gerchberg-Saxton algorithm and angular-spectrum layer-oriented method for true color three-dimensional display. Acta Physica Sinica, 2018, 67(9): 094203. doi: 10.7498/aps.67.20172464 [2] Zhang Dong-Xiao, Chen Zhi-Bin, Xiao Cheng, Qin Meng-Ze, Wu Hao. Generation of turbulence phase screen based on gravitational search algorithm. Acta Physica Sinica, 2019, 68(13): 134205. doi: 10.7498/aps.68.20190081 [3] MaYan-Xing, Wang Xiao-Lin, Zhou Pu, Ma Hao-Tong, Zhao Hai-Chuan, Xu Xiao-Jun, Si Lei, Liu Ze-Jin, Zhao Yi-Jun. Effect of atmosphere turbulence on phase modulation signals in coherent beam combination with multi-dithering technique. Acta Physica Sinica, 2011, 60(9): 094211. doi: 10.7498/aps.60.094211 [4] Cai Dong-Mei, Wang Kun, Jia Peng, Wang Dong, Liu Jian-Xia. Sampling methods of power spectral density method simulating atmospheric turbulence phase screen. Acta Physica Sinica, 2014, 63(10): 104217. doi: 10.7498/aps.63.104217 [5] Cai Dong-Mei, Ti Pei-Pei, Jia Peng, Wang Dong, Liu Jian-Xia. Fast simulation of atmospheric turbulence phase screen based on non-uniform sampling. Acta Physica Sinica, 2015, 64(22): 224217. doi: 10.7498/aps.64.224217 [6] Zheng Wei-Wei, Wang Li-Qin, Xu Jing-Ping, Wang Li-Gang. Studies on propagation of laser beam array with initial phase distributions in a turbulent atmosphere. Acta Physica Sinica, 2009, 58(7): 5098-5103. doi: 10.7498/aps.58.5098 [7] Zhang Yu-Yan, Zhou Hang, Yan Meisu. Study on the phase-extracting method of self-mixing signal based on empirical mode decomposition. Acta Physica Sinica, 2015, 64(5): 054203. doi: 10.7498/aps.64.054203 [8] Li Xiao-Qing, Ji Xiao-Ling, Zhu Jian-Hua. Higher-order intensity moments of optical beams in atmospheric turbulence. Acta Physica Sinica, 2013, 62(4): 044217. doi: 10.7498/aps.62.044217 [9] Chen Xiao-Wen, Tang Ming-Yue, Ji Xiao-Ling. The influence of atmospheric turbulence on the spatial correlation property of partially coherent Hermite-Gaussian beams. Acta Physica Sinica, 2008, 57(4): 2607-2613. doi: 10.7498/aps.57.2607 [10] Ji Xiao-Ling. Influence of atmospheric turbulence on the spreading and directionality of radial Gaussian array beams. Acta Physica Sinica, 2010, 59(1): 692-698. doi: 10.7498/aps.59.692
•  Citation:
##### Metrics
• Abstract views:  84
• Cited By: 0
##### Publishing process
• Received Date:  28 June 2019
• Accepted Date:  15 October 2019
• Published Online:  01 January 2020

## Extracting atmospheric turbulence phase using deep convolutional neural network

###### Corresponding author: Liu Jun-Min, liujunmin@sztu.edu.cn;
• 1. College of New Materials and New Energies, Shenzhen Technology University, Shenzhen 518118, China
• 2. Engineering Technology Research Center for 2D Material Information Function Devices and Systems of Guangdong Province, International Collaborative Laboratory of 2D Materials for Optoelectronics Science and Technology, Shenzhen University, Shenzhen 518060, China
• 3. Synergetic Innovation Center for Quantum Effects and Applications, School of Physics and Electronics, Hunan Normal University, Changsha 410081, China

Abstract: When a light beam transmits in free space, it is easily affected by atmospheric turbulence. The effect on transmitted light is equivalent to adding a random noise phase to it, which leads its transmission quality to deteriorate. The method of improving the quality of transmitted beams is usually to compensate for the phase distortion at the receiver by adding reverse turbulence phase, and the premise of this method is to obtain the turbulence phase carried by the distorted beam. The adaptive optics system is the most common way to extract the phase information. However, it is inefficient to be applied to varying turbulence environments due to the fact that a wave-front sensor and complex optical system are usually contained. Deep convolutional neural network (CNN) that can directly capture feature information from images is widely used in computer vision, language processing, optical information processing, etc. Therefore, in this paper proposed is a turbulence phase information extraction scheme based on the CNN, which can quickly and accurately extract the turbulence phase from the intensity patterns affected by atmosphere turbulence. The CNN model in this paper consists of 17 layers, including convolutional layers, pooling layers and deconvolutional layers. The convolutional layers and pooling layers are used to extract the turbulent phase from the feature image, which is the core structure of the network. The function of the deconvolutional layers is to visualize the extracted turbulence information and output the final predicted turbulence phase. After learning a huge number of samples, the loss function value of CNN converges to about 0.02, and the average loss function value on the test set is lower than 0.03. The trained CNN model has a good generalization capability and can directly extract the turbulent phase according to the input light intensity pattern. Using an I5-8500 CPU, the average time to predict the turbulent phase is as low as s under the condition of $C_{{n}}^2 = 1 \times {10^{ - 14}}\;{{\rm{m}}^{ - 2/3}}$, $5 \times {10^{ - 14}}\;{{\rm{m}}^{ - 2/3}}$, and $1 \times {10^{ - 13}}\;{{\rm{m}}^{ - 2/3}}$. In addition, the turbulence phase extraction capability of CNN can be further enhanced by improving computing power or optimizing model structure. These results indicate that the CNN-based turbulence phase extraction method can effectively extract the turbulence phase, which has important application value in turbulence compensation, atmospheric turbulence characteristics research and image reconstruction.

Reference (27)

/