-
关联成像(Correlated imaging),也被称作鬼成像(Ghost imaging),是一种基于光场高阶关联获取物体信息的成像技术。如图3所示,在关联成像系统中,光源被分为两路:一路经过物体后的总光强被一个点探测器收集(孔径内能量被透镜收集至一点,这种探测方式也称桶探测),被称为物臂;另一路是参考臂,经自由传播后的光强分布被面阵探测器记录。物体的图像可以通过两臂信号的关联计算得到。
与直接记录物体信号光强度分布的传统成像相比,关联成像有以下显著特点。首先,桶探测的方式实现了对回波能量的汇聚收集,大幅提升了信号的能量密度,同时单元探测器件在技术上往往比面阵探测器件具备更优性能,这使得关联成像可以在更弱回波条件下工作[14-16]。其次,关联成像通过照明或后端调制对物体图像信息进行编码,可根据成像所需设计编码方式,以实现成像所需能量、时间、数据量和采样次数的大幅减少[17-23]。再次,关联图像是通过照明编码信息和桶探测信号重构的,可以有效抑制与照明光源统计无关的噪声[24-28],因此可以大幅提升噪声条件下尤其是弱光条件下成像的鲁棒性。最后,关联成像可以结合压缩感知、机器学习等算法相结合,突破奈奎斯特采样定律,实现亚采样下的图像重构。而在照明功率恒定的条件下,大幅减少成像所需采样次数也意味着大幅减少了成像所需的照明能量。因此,上述特点共同形成了关联成像更高的灵敏度和在极弱光条件下的潜在优势。此外,关联成像还可以实现无透镜成像,可拓展至在X光、中、远红外、太赫兹波等一些难以实现面阵探测或难以获取成像透镜的波段。另一方面,由于使用了单元探测器,关联成像需多次测量来获取二维、三维乃至更高维的场景信息。
在关联成像系统中,“点到斑”的共轭成像关系是通过光场的高阶关联建立的。图4为关联成像系统的展开图,即将图3中的光路在分束器处展开。为了简洁起见同时不影响讨论,这里省略了探测器,并且以赝热光关联成像为例。根据菲涅尔衍射公式,可以得到物体表面的光场为[29]:
$$\begin{array}{l} {{\vec E}_s}\left( {{x_s},{y_s}} \right) = \int {{{\vec E}_0}} \left( {{x_0},{y_0}} \right){h_s}\left( {{x_0},{y_0};{x_s},{y_s};u} \right)\\ \;\;\;\;\;\;\;\;\;\;\;{\rm{d}}{x_0}{\rm{d}}{y_0},s = \{ r,t\} ,z = \{ v,u\} \end{array}$$ (1) 当
$s=r(t)$ 时表示参考臂(物臂),$\vec{E}_{0}\left(x_{0}, y_{0}\right) $ 表示光源表面光场,$ z=v(u) $ 为参考臂相机(物体)到光源的距离,$ h_{s}\left(x_{0}, y_{0} ; x_{s}, y_{s} ; u\right) $ 表示脉冲相应函数:$$\begin{split} {h_s}\left( {{x_0},{y_0};{x_s},{y_s};z} \right) &= \dfrac{{{{\rm{e}}^{ikz}}}}{{i\lambda z}}\;\exp \;\left\{ {\dfrac{{ik}}{{2z}}\left[ {{{\left( {{x_s} - {x_0}} \right)}^2} + {{\left( {{y_s} - {y_0}} \right)}^2}} \right]} \right\}\\ s &= \{ r,t\} ,z = \{ v,u\} \end{split}$$ (2) 式中:λ为波长; k为对应的角波数; i为虚数单位。根据复高斯矩定理[30],物体表面和参考臂相机表面光场的二阶关联函数可以写为:
$$\begin{split} &G\left( {{x_t},{y_t};{x_r},{y_r}} \right) = \left\langle {I\left( {{x_t},{y_t}} \right)I\left( {{x_r},{y_r}} \right)} \right\rangle = \\ &\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\left\langle {I\left( {{x_t},{y_t}} \right)} \right\rangle \left\langle {I\left( {{x_r},{y_r}} \right)} \right\rangle + {\left| {\left\langle {{{\vec E}_t}\left( {{x_t},{y_t}} \right)\vec E_r^*\left( {{x_r},{y_r}} \right)} \right\rangle } \right|^2} \end{split}$$ (3) 式中:〈·〉表示时域上的系综平均,实际中通过大量采样求平均值来实现;*表示光场的共轭。公式(3)中等号右边的第一项为常数项,不影响对图像信息传递的讨论。将公式(1)和(2)代入第二项,得到:
$$\begin{split} &\left|\left\langle\vec{E}_{t}\left(x_{t}, y_{t}\right) \vec{E}_{r}^{*}\left(x_{r}, y_{r}\right)\right\rangle\right|^{2}= \\ &\mid \frac{{\rm{e}}^{i k(u-v)}}{\lambda^{2} u^{2} v^{2}} \int\left\langle\vec{E}_{0}^{*}\left(x_{0}, y_{0}\right) \vec{E}_{0}^{\prime}\left(x_{0}^{\prime}, y_{0}^{\prime}\right)\right\rangle \cdot \\ &\exp \left\{\frac{-i k}{2 v}\left[\left(x_{r}-x_{0}\right)^{2}+\left(y_{r}-y_{0}\right)^{2}\right]\right\}\cdot \\ &\exp \left\{\frac { i k } { 2 u } \left[\left(x_{t}-x_{0}^{\prime}\right)^{2}+\left(y_{t}-\right.\right.\right. \\ &\left.\left.y_{0}^{\prime}\right)^{2}\right\} \mathrm{d} x_{0} \mathrm{~d} y_{0} \mathrm{~d} x_{0}^{\prime} \mathrm{d} y_{0}^{\prime} \mid \end{split}$$ (4) 对于完全非相干照明光源,有:
$$\left\langle {\vec E_0^*\left( {{x_0},{y_0}} \right)\vec E_0^\prime \left( {x_0^\prime ,y_0^\prime } \right)} \right\rangle = \left\langle {I\left( {{x_0},{y_0}} \right)} \right\rangle \delta \left( {{x_0} - x_0^\prime ,{y_0} - y_0^\prime } \right)$$ (5) 式中:
$\delta \left( \cdot \right) $ 为Dirac-delta函数;$ I\left(x_{0}, y_{0}\right) $ 为光源的强度分布。当参考臂和物臂光程相等,即$ v = u = d $ 时,将公式(5)代入公式(4),可以得到:$$\begin{split} \begin{array}{l} {\left| {\left\langle {{{\vec E}_t}\left( {{x_t},{y_t}} \right)\vec E_r^*\left( {{x_r},{y_r}} \right)} \right\rangle } \right|^2} = \alpha \left| {\int {\left\langle {I\left( {{x_0},{y_0}} \right)} \right\rangle } } \right.\cdot \\ {\left. {\exp \left\{ {\dfrac{{ - i2\pi }}{{\lambda d}}\left[ {\left( {{x_r} - {x_t}} \right){x_0} + \left( {{y_r} - {y_t}} \right){y_0}} \right]} \right\}{\rm{d}}{x_0}{\rm{d}}{y_0}} \right|^2} = \end{array}\\ {\;\;\;\;\;\;\;\;\;\alpha {{\left| {F\left[ {I\left( {\dfrac{{{x_r} - {x_t}}}{{\lambda d}},\dfrac{{{y_r} - {y_t}}}{{\lambda d}}} \right)} \right]} \right|}^2}} \end{split}$$ (6) 式中:
$ \alpha = \dfrac{1}{{{\lambda ^2}{d^4}}} $ ;${F}[ \cdot ]$ 表示傅里叶变换。因此,忽略公式(3)的常数项时,物体表面和参考臂相机表面光场的二阶关联为光源强度分布的傅里叶变换的模平方。对于旋转毛玻璃调制高斯激光束产生的赝热光[31-35],光源强度服从高斯分布,其傅里叶变换也服从高斯分布,且当$ x_{r} \cong x_{t} $ &$ y_{r} \cong y_{t} $ ,时出现最值。该高斯分布的半高宽取决于光源孔径,光源孔径越大,半高宽越小,反之半高宽越大。这表明,照在物体上(xt, yt)的光场,和参考臂相机表面以(xr, yr)为中心的微小区域(区域面积取决于光源孔径)光场的关联性最强,和其他区域的光场无关联。当桶探测器收集到来自(xt, yt)的光强并和参考臂记录的光场进行关联运算时,在(xr, yr)点的关联信号就会明显强于其他点。这就建立了关联成像系统中“点到斑”的成像关系。并且,由于光场关联的条件是$ x_{r} \cong x_{t} $ &$ y_{r} \cong y_{t} $ ,因此关联成像系统得到的是正立的图像。除了上述的强度关联理论,马里兰大学的史砚华教授也用双光子干涉理论解释了关联成像的物理机制[36-38]。在极弱光环境下,关联成像质量也会降低,但像质受影响的机理和传统成像具有显著差异。以赝热光关联成像为例,当照在物体上的光场或物体的回波弱至少光子乃至单光子时,由于光子数噪声、环境噪声以及探测器的影响,导致桶探测信号和参考臂共轭点的强度统计关联下降,而和非共轭点的统计关联性则有可能上升,进而影响成像中“点到斑”的共轭关系。对于纠缠光关联成像[39-41],在极弱光环境下,虽然纠缠光子对自身的关联性并不会降低,但在背景噪声和非理想器件的影响下,实际探测到两臂信号的关联性也会下降,因此成像质量也会退化[42-43]。
近年来,研究人员为提升极弱光环境下关联成像的性能进行了广泛深入的研究,并取得了诸多进展。包括大幅度降低成像所需辐射剂量[44-45],在极低探测信噪比下实现图像重构[46-49],在平均每像素的光子数远小于1的极弱光条件下获得物体图像[50-53]等。
Progress and prospect of ghost imaging in extremely weak light (Invited)
-
摘要: 光学成像因其分辨率高,信息量丰富,具有其他探测和感知技术不可替代的地位,是人们获取信息最重要的技术手段之一。光子是光学成像系统中的信息载体。光学图像的高质量重构,依赖于对信号光子的高效耦合和对光信息的精准解耦。然而,在遥感或生物成像等重要应用场景中,由于作用距离远或辐照功率低,到达探测面的物体信号光子数少,信噪比低,对光学系统设计、信号探测和图像恢复都带来了极大困难,严重限制了光学成像性能。如何在极弱光条件下获得高质量图像,是光电成像系统研究的基础性难题,也是推动光学成像不断向更大视场、更远作用距离、更高信息通量发展亟待克服的关键技术。近年来,在光场调控和量子探测技术支撑下,并基于光场的高阶经典/量子关联发展起来的关联成像,由于探测灵敏度高、抗干扰能力强,为发展极弱光条件下的光学成像技术带来了新的机遇。文中将简要回顾关联成像的原理机制,在此基础上系统介绍极弱光条件下关联成像方案和方法。并尝试从光子动力学层面解释这些方法的物理本质,讨论这些方法的能力极限,比较这些方法所适用的场景。Abstract: Possessing high resolution and rich information, optical imaging is one of the most important techniques for people to obtain information. Photons are information carriers in optical imaging systems. The high-quality reconstruction of optical image depends on the efficient coupling of signal photons and the accurate decoupling of optical information. However, in important application scenarios such as remote sensing or biological imaging, due to the long operating distance or low radiation power, the number of signal photons from the object to the detection plane is small, and the signal-to-noise ratio is low, thus bringing great difficulties to the design of optical system, the signal detection and the image reconstruction, and seriously limiting the performance of optical imaging. How to obtain high-quality images under extremely weak light conditions is not only a basic problem of photoelectric imaging system research, but also a key technology to promote the vigorous development of optical imaging with a larger field of view, longer working distance and higher information flux. In recent years, with the support of light field modulation and quantum detection technology, and based on the high-order classical/quantum correlation of light field, ghost imaging has brought new opportunities for the development of optical imaging technology under extremely weak light conditions, due to its high detection sensitivity and strong ability against interference. This paper briefly reviewed the principle and mechanism of ghost imaging, and systematically introduced the schemes and methods of ghost imaging under very weak light conditions. The physical essence of these methods from the level of photon dynamics was introduced, the capability limits of these methods were discussed, and the applicable scenarios of these methods were compared.
-
Key words:
- imaging system /
- ghost imaging /
- weak light imaging /
- photon counting /
- single photon imaging
-
图 2 (a)首达光子成像实验装置图;(b)首达光子成像实验结果:第一列为雪崩二极管记录的点云图,第二列为通过首达光子飞行时间估计得到的反射率,第三列为正则化计算得到的图像,第四列为根据不同视图估计得到的三维图像[12]
Figure 2. (a) Experimental setup of first-photon imaging; (b) Experimental results of first-photon imaging: The first column is the point cloud recorded by the avalanche photodiode, the second column is the reflectivity estimated from the time-of-flight of the first photon, the third column is the computational image via regularization method, the fourth column is the 3-dimentianl image estimated from the different views[12]
图 10 (a) 1.3 km关联成像示意图;(b)时域关联方法和传统方法的实验结果比较; (c)时域关联方法对不同距离上的物体成像结果[49]
Figure 10. (a) Diagrammatic sketch of ghost imaging at 1.3 km; (b) Comparison between the experimental results of the temporal correlation method and that from traditional method; (c) Imaging results of objects at different distance from the temporal correlation method[49]
-
[1] Lichtman J W, Conchello J A. Fluorescence microscopy [J]. Nature Methods, 2005, 2(12): 910-919. doi: 10.1038/nmeth817 [2] Gardner J P, Mather J C, Clampin M, et al. The james webb space telescope [J]. Space Science Reviews, 2006, 123(4): 485-606. doi: 10.1007/s11214-006-8315-7 [3] Kobayashi M, Kikuchi D, Okamura H. Imaging of ultraweak spontaneous photon emission from human body displaying diurnal rhythm [J]. PLoS One, 2009, 4(7): e6256. doi: 10.1371/journal.pone.0006256 [4] Cao W, Che R, Ye D. An illumination-independent edge detection and fuzzy enhancement algorithm based on wavelet transform for non-uniform weak illumination images [J]. Pattern Recognition Letters, 2008, 29(3): 192-199. doi: 10.1016/j.patrec.2007.09.012 [5] Yeom S, Javidi B, Watson E. Photon counting passive 3 D image sensing for automatic target recognition [J]. Optics Express, 2005, 13(23): 9310-9330. doi: 10.1364/OPEX.13.009310 [6] Johnson S D, Moreau P A, Gregory T, et al. How many photons does it take to form an image? [J]. Applied Physics Letters, 2020, 116(26): 260504. doi: 10.1063/5.0009493 [7] You L, Yang X, He Y, et al. Jitter analysis of a superconducting nanowire single photon detector [J]. Aip Advances, 2013, 3(7): 072135. doi: 10.1063/1.4817581 [8] You L. Superconducting nanowire single-photon detectors for quantum information [J]. Nanophotonics, 2020, 9(9): 2673-2692. doi: 10.1515/nanoph-2020-0186 [9] Zhou Hui, Zhang Chengjun, Lü Chaolin, et al. Recent progress of imaging applications based on superconducting nanowire single-photon detectors [J]. Laser & Optoelectronics Progress, 2021, 58(10): 1011005. (in Chinese) [10] Kong Lingdong, Zhao Qingyuan, Tu Xuecou, et al. Progress and applications of superconducting nanowire delay-line single-photon imagers [J]. Laser & Optoelectronics Progress, 2021, 58(10): 1011002. (in Chinese) [11] Sciamanda R J. Dirac and photon interference [J]. American Journal of Physics, 1969, 37(11): 1128-1130. doi: 10.1119/1.1975229 [12] Kirmani A, Venkatraman D, Shin D, et al. First-photon imaging [J]. Science, 2014, 343(6166): 58-61. [13] Li Zhengping, Ye Juntian, Xin Huang, et al. Single-photon imaging over 200 km [J]. Optica, 2021, 8(3): 344-349. doi: 10.1364/OPTICA.408657 [14] Morris P A, Aspden R S, Bell J E C, et al. Imaging with a small number of photons [J]. Nature Communications, 2015, 6: 5913. doi: 10.1038/ncomms6913 [15] Aspden R S, Tasca D S, Boyd R W, et al. EPR-based ghost imaging using a single-photon-sensitive camera [J]. New Journal of Physics, 2013, 15(7): 073032. doi: 10.1088/1367-2630/15/7/073032 [16] Zhu Y, Shi J H, Wu X Y, et al. Photon-limited non-imaging object detection and classification based on single-pixel imaging system [J]. Appl Phys B, 2020, 126: 21. [17] Ma S, Hu C Y, Wang C L, et al. Multi-scale ghost imaging LiDAR via sparsity constraints using push-broom scanning [J]. Opt Communication, 2019, 448: 89-92. [18] Chen M L, Li E R, Han S S. Application of multi-correlation-scale measurement matrices in ghost imaging via sparsity constraints [J]. App Opt, 2014, 53: 13. [19] Aβmann M, Bayer M. Compressive adaptive computational ghost imaging [J]. Scientific Reports, 2013, 3: 1545. doi: 10.1038/srep01545 [20] Yu W K, Li M F, Yao X R, et al. Adaptive compressive ghost imaging based on wavelet trees and sparse representation [J]. Optics Express, 2014, 22(6): 7133-7144. doi: 10.1364/OE.22.007133 [21] Gu J H, Sun S, Xu Y K, et al. Feedback ghost imaging by gradually distinguishing and concentrating onto the edge area [J]. Chin Opt Lett, 2021, 19(2): 0411. [22] Sun S, Liu W T, Lin H Z, et al. Multi-scale adaptive computational ghost imaging [J]. Scientific Reports, 2016, 6: 37013. doi: 10.1038/srep37013 [23] Sun S, Lin H Z, Xu Y K, et al. Tracking and imaging of moving objects with temporal intensity difference correlation [J]. Optics Express, 2019, 27(20): 27851-27861. doi: 10.1364/OE.27.027851 [24] Xu Y K, Liu W T, Zhang E F, et al. Is ghost imaging intrinsically more powerful against scattering? [J]. Opti Express, 2015, 23(26): 32993-33000. doi: 10.1364/OE.23.032993 [25] Wu Yongbo, Yang Zhihui, Tang Zhilie. Experimental study on anti-disturbance ability of underwater ghost imaging [J]. Laser & Optoelectronics Progress, 2021, 58(6): 0611002. (in Chinese) [26] Liu Weitao, Sun Shuai, Hu Hongkang, et al. Progress and prospect for ghost imaging of moving objects [J]. Laser & Optoelectronics Progress, 2021, 58(10): 1011001. (in Chinese) [27] Gong Wenlin, Sun Jianfeng, Deng Chenjin, et al. Research progress on single-pixel imaging lidar via coherent detection [J]. Laser & Optoelectronics Progress, 2021, 58(10): 1011003. (in Chinese) [28] Fu Xiquan, Huang Xianwei, Tan Wei, et al. Correlation imaging research under disturbance of channel airflow [J]. Laser & Optoelectronics Progress, 2021, 58(10): 1011017. (in Chinese) [29] Goodman, Joseph W. Statistical Optics[M]. New York: John Wiley & Sons, 2015. [30] Gatti A, Bache M, Magatti D, et al. Coherent imaging with pseudo-thermal incoherent light [J]. Journal of Modern Optics, 2006, 53(5-6): 739-760. [31] Shapiro J H. Computational ghost imaging [J]. Physical Review A, 2008, 78(6): 061802. doi: 10.1103/PhysRevA.78.061802 [32] Hardy N D, Shapiro J H. Computational ghost imaging versus imaging laser radar for three-dimensional imaging [J]. Physical Review A, 2013, 87(2): 023820. doi: 10.1103/PhysRevA.87.023820 [33] Liu X F, Chen X H, Yao X R, et al. Lensless ghost imaging with sunlight [J]. Opt Lett, 2014, 39(8): 2314-2317. doi: 10.1364/OL.39.002314 [34] Gong W, Zhao C, Yu H, et al. Three-dimensional ghost imaging lidar via sparsity constraint [J]. Scientific Reports, 2016, 6: 26133. doi: 10.1038/srep26133 [35] Gong W, Han S. Correlated imaging in scattering media [J]. Optics Letters, 2011, 36(3): 394-396. doi: 10.1364/OL.36.000394 [36] Shih Y. Classical, Semi-classical and Quantum Noise[M].Berlin: Springs, 2012: 169-222. [37] Shih Y. The physics of ghost imaging: nonlocal interference or local intensity fluctuation correlation? [J]. Quantum Information Processing, 2012, 11(4): 995-1001. doi: 10.1007/s11128-012-0396-5 [38] Shih Y. The physics of turbulence-free ghost imaging [J]. Technologies, 2016, 4(4): 39. doi: 10.3390/technologies4040039 [39] Klyshko D N. Photon and Nonlinear Optics [M]. New York: Gordon and Breach Science Press, 1988. [40] Bornman N, Agnew M, Zhu F, et al. Ghost imaging using entanglement-swapped photons [J]. Quantum Information, 2019, 5(1): 1-6. doi: 10.1038/s41534-018-0113-z [41] Zerom P, Chan K W C, Howell J C, et al. Entangled-photon compressive ghost imaging [J]. Physical Review A, 2011, 84(6): 061804. doi: 10.1103/PhysRevA.84.061804 [42] Li J, Gao W, Qian J, et al. Robust entangled-photon ghost imaging with compressive sensing [J]. Sensors, 2019, 19(1): 192. doi: 10.3390/s19010192 [43] Dixon P B, Howland G A, Chan K W C, et al. Quantum ghost imaging through turbulence [J]. Physical Review A, 2011, 83(5): 051803. doi: 10.1103/PhysRevA.83.051803 [44] Schori A, Shwartz S. X-ray ghost imaging with a laboratory source [J]. Optics Express, 2017, 25(13): 14822-14828. doi: 10.1364/OE.25.014822 [45] Zhang A X, He Y H, Wu L A, et al. Tabletop x-ray ghost imaging with ultra-low radiation [J]. Optica, 2018, 5(4): 374-377. doi: 10.1364/OPTICA.5.000374 [46] Deng Chenjin, Gong Wenlin, Han Shensheng. Pulse-compression ghost imaging lidar via coherent detection. [J]. Optics Express, 2016, 24(23): 25983-25994. doi: 10.1364/OE.24.025983 [47] Deng Chenjin, Pan Long, Wang Chenglong, et al. Performance analysis of ghost imaging lidar in background light environment [J]. Photonics Research, 2017(5): 431-435. [48] Pan Long, Deng Chenjin, Bo Zunwang, et al. Experimental investigation of chirped amplitude modulation heterodyne ghost imaging. [J]. Optics Express, 2020, 28(14): 20808-20816. doi: 10.1364/OE.399006 [49] Li Dong, Yang Ding, Sun Shuai, et al. Enhancing robustness of ghost imaging against environment noise via cross-correlation in time domain. [J]. Optics Express, 2021, 29(20): 31068-31077. doi: 10.1364/OE.439519 [50] Yang Ying, Shi Jianhong, Cao Fei, et al. Computational imaging based on time-correlated single-photon-counting technique at low light level [J]. Applied Optics, 2015, 54(31): 009277. [51] Liu Y, Shi J, Zeng G. Single-photon-counting polarization ghost imaging [J]. Applied Optics, 2016, 55(36): 10347. doi: 10.1364/AO.55.010347 [52] Liu X, Shi J, Wu X, et al. Fast first-photon ghost imaging [J]. Scientific Reports, 2018, 8(1): 5012. doi: 10.1038/s41598-018-23363-w [53] Liu X, Shi J, Sun L, et al. Photon-limited single-pixel imaging [J]. Optics Express, 2020, 28(6): 8132. doi: 10.1364/OE.381785 [54] Pittman T B, Shih Y H, Strekalov D V, et al. Optical imaging by means of two-photon quantum entanglement [J]. Physical Review A, 1995, 52: R3429. doi: 10.1103/PhysRevA.52.R3429 [55] Aspden R S, Morris P A, He R, et al. Heralded phase-contrast imaging using an orbital angular momentum phase-filter [J]. Journal of Optics, 2016, 18(5): 055204. doi: 10.1088/2040-8978/18/5/055204 [56] Tasca D S, Aspden R S, Morris P A, et al. The influence of non-imaging detector design on heralded ghost-imaging and ghost-diffraction examined using a triggered ICCD camera [J]. Optics Express, 2013, 21(25): 30460-30473. doi: 10.1364/OE.21.030460 [57] Liu Shikai, Zhou Zhiyuan, Shi Baosen. Progress on optical image edge detection [J]. Laser & Optoelectronics Progress, 2021, 58(10): 1011014. (in Chinese) [58] Brida G, Genovese M, Berchera I R. Experimental realization of sub-shot-noise quantum imaging [J]. Nature Photonics, 2010, 4(4): 227. doi: 10.1038/nphoton.2010.29 [59] Genovese M. Real applications of quantum imaging [J]. Journal of Optics, 2016, 18(7): 073002. doi: 10.1088/2040-8978/18/7/073002 [60] Samantaray N, Ruo-Berchera I, Meda A, et al. Realization of the first sub-shot-noise wide field microscope [J]. Light: Science & Applications, 2017, 6(7): e17005. [61] Valencia Alejandra, Scarcelli Giuliano, D'Angelo Milena, et al. Two-photon imaging with thermal light. [J]. Physical Review Letters, 2005, 94(6): 063601. doi: 10.1103/PhysRevLett.94.063601 [62] Yang Dongyue, Wu Guohua, Li Junhui, et al. Image recovery of ghost imaging with sparse spatial frequencies [J]. Optics Letters, 2020, 45(19): 403288. [63] Sun Shuai, Liu Weitao, Gu Junhao, et al. Ghost imaging normalized by second-order coherence. [J]. Optics Letters, 2019, 44(24): 5993-5996. doi: 10.1364/OL.44.005993 [64] Wang Fei, Wang Hao, Wang Haichao, et al. Learning from simulation: An end-to-end deep-learning approach for computational ghost imaging. [J]. Optics Express, 2019, 27(18): 25560-25572. doi: 10.1364/OE.27.025560 [65] Hu Hongkang, Sun Shuai, Lin Huizu, et al. Denoising ghost imaging under a small sampling rate via deep learning for tracking and imaging moving objects. [J]. Optics Express, 2020, 28(25): 37284-37293. doi: 10.1364/OE.412597 [66] He Y C, Wang G, Dong G X. et al. Ghost imaging based on deep learning [J]. Sci Rep, 2018, 8: 6469. [67] Wang F, Wang H, Wang H C, et al. Learning from simulation: An end-to-end deep-learning approach for computational ghost imaging [J]. Opt Express, 2019, 27: 25560-25572. [68] Rizvi S, Cao J, Zhang K Y, et al. DeepGhost: real-time computational ghost imaging via deep learning [J]. Sci Rep, 2020, 10: 1140. [69] Yang Y, Shi J, Cao F, et al. Computational imaging based on time-correlated single-photon-counting technique at low light level [J]. Applied Optics, 2015, 54(31): 9277-9283. [70] Liu H C, Yang H, Xiong J, et al. Positive and negative ghost imaging [J]. Physical Review Applied, 2019, 12(3): 034019. doi: 10.1103/PhysRevApplied.12.034019 [71] Li G L, Zhao Y, Yang Z H, et al. Positive–negative corresponding normalized ghost imaging based on an adaptive threshold [J]. Laser Physics Letters, 2016, 13(11): 115202. doi: 10.1088/1612-2011/13/11/115202 [72] Yang H, Wu S, Wang H B, et al. Probability theory in conditional-averaging ghost imaging with thermal light [J]. Physical Review A, 2018, 98(5): 053853. doi: 10.1103/PhysRevA.98.053853 [73] Cao D Z, Zhang S H, Zhao Y, et al. Zero-photon imaging under extremely low-light illumination [J]. arXiv preprint, 2021, 2108: 01037.