基于数据驱动的发射端波前校正鬼成像系统*

doi:10.7523/j.ucas.2026.038

摘要/Abstract

摘要： 鬼成像技术在生物医学、遥感探测与水下通信等领域极具应用潜力。该技术基于二阶相关成像的统计特性,对接收路径上的散射干扰具有良好的抗性。然而,若照明路径出现波前散射畸变,散斑的关联性会被严重破坏,导致成像质量骤降。现有解决方案多依赖手工设计投影图案或进行复杂的物理建模。这些方法泛化能力弱、效率低,难以应对多样化的静态畸变环境。为此,本研究提出了重度散射干扰的鬼成像波前预补偿方法,设计了一套基于发射端波前优化闭环反馈的鬼成像系统,以期改善重度散射干扰下的成像质量。该系统引入深度学习技术,设计了反馈调节散斑畸变网络（FRSD-Net）,通过闭环反馈机制智能生成最优照明散斑,有效保证了光场在穿透重度非均匀介质后,仍能保持高信噪比与强关联特性。实验表明,在水渍及聚合物等重度畸变场景下,该系统显著提升了鬼成像的质量。图像的PSNR/SSIM分别从6.45/0.12提升至22.00/0.73,以及从15.71/0.32提升至23.19/0.68。该方案有效结合了DMD的高速调制优势与深度学习的建模能力,为鬼成像在复杂散射环境中的实际应用提供了新路径。

关键词: 鬼成像, 波前整形, 数字微镜器件, 深度学习, 散射补偿

Abstract: Ghost imaging (GI) technology holds significant application potential in fields such as biomedicine, remote sensing, and underwater communication. Based on the statistical characteristics of second-order correlation imaging, this technology exhibits strong robustness against scattering interference in the receiving path. However, when wavefront scattering distortion occurs in the illumination path, the speckle correlation is severely degraded, leading to a drastic decline in imaging quality. Existing solutions mostly rely on manually designed projection patterns or complex physical modeling. These methods suffer from weak generalization ability and low efficiency, making it difficult to cope with diverse static distortion environments. To address this, this study proposes a wavefront pre-compensation method for ghost imaging under severe scattering interference and designs a ghost imaging system based on closed-loop feedback of transmitter-end wavefront optimization. By introducing deep learning techniques, the system incorporates a Feedback-Regulated Speckle Distortion Network (FRSD-Net). Through a closed-loop feedback mechanism, it intelligently generates optimal illumination speckles, effectively ensuring that the light field maintains a high signal-to-noise ratio (SNR) and strong correlation characteristics even after penetrating severely inhomogeneous media. Experimental results demonstrate that the proposed system significantly improves ghost imaging quality under severe distortion scenarios, such as those caused by water stains and polymers. The PSNR/SSIM of the reconstructed images increased from 6.45/0.12 to 22.00/0.73, and from 15.71/0.32 to 23.19/0.68, respectively. This approach effectively combines the high-speed modulation advantages of Digital Micromirror Devices (DMD) with the modeling capabilities of deep learning, providing a new pathway for the practical application of ghost imaging in complex scattering environments.

Key words: Ghost imaging, Wavefront shaping, Digital micromirror device, Deep learning, Scattering compensation

中图分类号:

O436

褚博文, 唐家奎, 郎森, 晏芳. 基于数据驱动的发射端波前校正鬼成像系统^*[J]. 中国科学院大学学报, DOI: 10.7523/j.ucas.2026.038.

Bowen Chu, Jiakui Tang, Sen Lang, Fang Yan. A data-driven Ghost Imaging System with transmitter-end wavefront correction[J]. Journal of University of Chinese Academy of Sciences, DOI: 10.7523/j.ucas.2026.038.

参考文献

[1] Ntziachristos V.Going deeper than microscopy: The optical imaging frontier in biology[J]. Nature Methods, 2010, 7(8): 603-614. DOI:10.1038/nmeth.1483.
[2] Gigan S.Optical microscopy aims deep[J]. Nature Photonics, 2017, 11(1): 14-16. DOI:10.1038/nphoton.2016.257.
[3] Bertolotti J, van Putten E G, Blum C, et al. Non-invasive imaging through opaque scattering layers[J]. Nature, 2012, 491(7423): 232-234. DOI:10.1038/nature11578.
[4] Katz O, Heidmann P, Fink M, et al.Non-invasive single-shot imaging through scattering layers and around corners via speckle correlations[J]. Nature Photonics, 2014, 8(10): 784-790. DOI:10.1038/nphoton.2014.189.
[5] Rotter S, Gigan S.Light fields in complex media: Mesoscopic scattering meets wave control[J]. Reviews of Modern Physics, 2017, 89: 015005. DOI:10.1103/revmodphys.89.015005.
[6] Mosk A P, Lagendijk A, Lerosey G, et al.Controlling waves in space and time for imaging and focusing in complex media[J]. Nature Photonics, 2012, 6(5): 283-292. DOI:10.1038/nphoton.2012.88.
[7] Popoff S M, Lerosey G, Carminati R, et al.Measuring the transmission matrix in optics: An approach to the study and control of light propagation in disordered media[J]. Physical Review Letters, 2010, 104(10): 100601. DOI:10.1103/physrevlett.104.100601.
[8] Horstmeyer R, Ruan H W, Yang C.Guidestar-assisted wavefront-shaping methods for focusing light into biological tissue[J]. Nature Photonics, 2015, 9(9): 563-571. DOI:10.1038/nphoton.2015.140.
[9] 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(5): R3429-R3432. DOI:10.1103/physreva.52.r3429.
[10] Shapiro J H.Computational ghost imaging[J]. Physical Review A, 2008, 78(6): 061802. DOI:10.1103/physreva.78.061802.
[11] Katz O, Bromberg Y, Silberberg Y.Compressive ghost imaging[J]. Applied Physics Letters, 2009, 95(13): 131110. DOI:10.1063/1.3238296.
[12] Erkmen B I, Shapiro J H.Ghost imaging: From quantum to classical to computational[J]. Advances in Optics and Photonics, 2010, 2(4): 405-450.
[13] Gibson G M, Johnson S D, Padgett M J.Single-pixel imaging 12 years on: A review[J]. Optics Express, 2020, 28(19): 28190. DOI:10.1364/oe.403195.
[14] Meyers R E, Deacon K S, Shih Y.Turbulence-free ghost imaging[J]. Applied Physics Letters, 2011, 98(11): 111115. DOI:10.1063/1.3567931.
[15] Tajahuerce E, Durán V, Clemente P, et al.Image transmission through dynamic scattering media by single-pixel photodetection[J]. Optics Express, 2014, 22(14): 16945. DOI:10.1364/oe.22.016945.
[16] Vellekoop I M, Mosk A P.Focusing coherent light through opaque strongly scattering media[J]. Optics Letters, 2007, 32(16): 2309. DOI:10.1364/ol.32.002309.
[17] Vellekoop I M.Feedback-based wavefront shaping[J]. Optics Express, 2015, 23(9): 12189. DOI:10.1364/oe.23.012189.
[18] Popoff S, Lerosey G, Fink M, et al.Image transmission through an opaque material[J]. Nature Communications, 2010, 1: 81. DOI:10.1038/ncomms1078.
[19] Judkewitz B, Wang Y M, Horstmeyer R, et al.Speckle-scale focusing in the diffusive regime with time reversal of variance-encoded light (TROVE)[J]. Nature Photonics, 2013, 7(4): 300-305. DOI:10.1038/nphoton.2013.31.
[20] Park J H, Yu Z P, Lee K, et al.Perspective: Wavefront shaping techniques for controlling multiple light scattering in biological tissues: Toward in vivo applications[J]. APL Photonics, 2018, 3(10): 100901. DOI:10.1063/1.5033917.
[21] Goorden S A, Bertolotti J, Mosk A P.Superpixel-based spatial amplitude and phase modulation using a digital micromirror device[J]. Optics Express, 2014, 22(15): 17999-18009. DOI:10.1364/OE.22.017999.
[22] Sun B, Edgar M P, Bowman R, et al.3D computational imaging with single-pixel detectors[J]. Science, 2013, 340(6134): 844-847. DOI:10.1126/science.1234454.
[23] Chen Y F, An H J, Sun Z, et al. Large model enhanced computational ghost imaging[EB/OL]. 2025: arXiv: 2503.08710. https://arxiv.org/abs/2503.08710
[24] Huang Y Y, Loesel P D, Paganin D M, et al. Deep learning in classical X-ray ghost imaging for dose reduction[EB/OL]. 2024: arXiv: 2411.06340. https://arxiv.org/abs/2411.06340
[25] Booth M J.Adaptive optical microscopy: The ongoing quest for a perfect image[J]. Light: Science & Applications, 2014, 3(4): e165. DOI:10.1038/lsa.2014.46.
[26] Ji N.Adaptive optical fluorescence microscopy[J]. Nature Methods, 2017, 14(4): 374-380. DOI:10.1038/nmeth.4218.
[27] Li Y G, Wang J J, Xiao T L, et al. Real-time imaging through dynamic scattering media enabled by fixed optical modulations[EB/OL]. 2025: arXiv: 2508.15286. https://arxiv.org/abs/2508.15286
[28] Du Y T, Liu Z C, Huang Y, et al. Fast physics-driven untrained network for highly nonlinear inverse scattering problems[EB/OL]. 2026: arXiv: 2602.13805. https://arxiv.org/abs/2602.13805
[29] Li Y G, Zheng Z, Wang J J, et al. Optical diffraction neural networks assisted computational ghost imaging through dynamic scattering media[EB/OL]. 2025: arXiv: 2511.22913. https://arxiv.org/abs/2511.22913
[30] González H, Martínez-León L, Soldevila F, et al. High sampling rate single-pixel digital holography system employing a DMD and phase-encoded patterns[EB/OL]. 2024: arXiv: 2402.03786. https://arxiv.org/abs/2402.03786
[31] Lee K, Park Y.Exploiting the speckle-correlation scattering matrix for a compact reference-free holographic image sensor[J]. Nature Communications, 2016, 7: 13359. DOI:10.1038/ncomms13359.
[32] Tzang O, Caravaca-Aguirre A M, Wagner K, et al. Adaptive wavefront shaping for controlling nonlinear multimode interactions in optical fibres[J]. Nature Photonics, 2018, 12(6): 368-374. DOI:10.1038/s41566-018-0167-7.
[33] Turtaev S, Leite I T, Mitchell K J, et al.Comparison of nematic liquid-crystal and DMD based spatial light modulation in complex photonics[J]. Optics Express, 2017, 25(24): 29874. DOI:10.1364/oe.25.029874.
[34] Barbastathis G, Ozcan A, Situ G H.On the use of deep learning for computational imaging[J]. Optica, 2019, 6(8): 921. DOI:10.1364/optica.6.000921.
[35] Manni M, Karpov D, Batenburg K J, et al. Noise2Ghost: Self-supervised deep convolutional reconstruction for ghost imaging[EB/OL]. 2025: arXiv: 2504.10288. https://arxiv.org/abs/2504.10288
[36] Tudosie S C, Gandolfi V, Varakkoth S, et al. Learned single-pixel fluorescence microscopy[EB/OL]. 2025: arXiv: 2507.18740. https://arxiv.org/abs/2507.18740
[37] Ferri F, Magatti D, Gatti A, et al.High-resolution ghost image and ghost diffraction experiments with thermal light[J]. Physical Review Letters, 2005, 94(18): 183602. DOI:10.1103/physrevlett.94.183602.
[38] Valencia A, Scarcelli G, D’Angelo M, et al. Two-photon imaging with thermal light[J]. Physical Review Letters, 2005, 94(6): 063601. DOI:10.1103/PhysRevLett.94.063601.
[39] D’Angelo M, Shih Y H. Quantum imaging[J]. Laser Physics Letters, 2005, 2(12): 567-596. DOI:10.1002/lapl.200510054.
[40] Bromberg Y, Katz O, Silberberg Y.Ghost imaging with a single detector[J]. Physical Review A, 2009, 79(5): 053840. DOI:10.1103/physreva.79.053840.
[41] Gong W L, Han S S.High-resolution far-field ghost imaging via sparsity constraint[J]. Scientific Reports, 2015, 5: 9280. DOI:10.1038/srep09280.
[42] Bina M, Magatti D, Molteni M, et al.Backscattering differential ghost imaging in turbid media[J]. Physical Review Letters, 2013, 110(8): 083901. DOI:10.1103/physrevlett.110.083901.
[43] Candes E J, Romberg J, Tao T.Robust uncertainty principles: Exact signal reconstruction from highly incomplete frequency information[J]. IEEE Transactions on Information Theory, 2006, 52(2): 489-509. DOI:10.1109/TIT.2005.862083.
[44] Donoho D L.Compressed sensing[J]. IEEE Transactions on Information Theory, 2006, 52(4): 1289-1306. DOI:10.1109/TIT.2006.871582.
[45] Erkmen B I, Shapiro J H.Signal-to-noise ratio of Gaussian-state ghost imaging[J]. Physical Review A, 2009, 79(2): 023833. DOI:10.1103/physreva.79.023833.

基于数据驱动的发射端波前校正鬼成像系统^*

A data-driven Ghost Imaging System with transmitter-end wavefront correction

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

访问统计

联系我们

[1]	姜翠, 焦建彬. 融合知识先验的车辆轨迹预测深度学习算法[J]. 中国科学院大学学报, 2026, 43(2): 209-217.
[2]	孙瑞奇, 张文娟, 李震, 马雪松, 梅君林. 分块尺寸对大场景遥感图像空间超分应用的影响分析[J]. 中国科学院大学学报, 2026, 43(1): 93-103.
[3]	周庆泽, 郭擎. 多重残差网络的多光谱遥感图像锐化方法[J]. 中国科学院大学学报, 2025, 42(4): 565-575.
[4]	曾建顺, 吕炎杰, 覃驭楚. 基于多尺度语义先验的街景图像修复算法[J]. 中国科学院大学学报, 2025, 42(4): 496-507.
[5]	钟金彦, 陈俊, 李宇, 吴业炜, 葛小青. 基于MFF-SFE的遥感图文跨模态检索方法[J]. 中国科学院大学学报, 2025, 42(2): 236-247.
[6]	周文雪, 张华春. 一种面向SAR图像快速舰船检测的轻量化网络[J]. 中国科学院大学学报, 2024, 41(6): 776-785.
[7]	陈经纬, 李宇, 陈俊, 张洪群. 基于MFF-Deeplabv3+网络的高分辨率遥感影像建筑物提取方法[J]. 中国科学院大学学报, 2024, 41(5): 654-664.
[8]	喻永生, 罗铁坚. 基于GAN反演的无缝图像补全技术[J]. 中国科学院大学学报, 2024, 41(5): 705-714.
[9]	黄玉林, 梁磊, 李卫军, 习晓环. 基于多尺度特征和注意力机制的深度学习点云压缩[J]. 中国科学院大学学报, 2024, 41(5): 687-694.
[10]	林昱彤, 王红, 柴团耀. 基于卷积神经网络多尺度特征的大豆基因组表型预测[J]. 中国科学院大学学报, 2024, 41(4): 468-476.
[11]	肖俊, 石光田. 三维点云去噪技术[J]. 中国科学院大学学报, 2023, 40(5): 577-595.
[12]	盛晓光, 王颖, 张迎伟, 项若曦, 付红萍. WILS：面向学业预警的非均衡增量式学习方法[J]. 中国科学院大学学报, 2023, 40(3): 422-432.
[13]	姚沐风, 昝露洋, 李柏鹏, 李庆亭, 陈正超. 基于CAR-Siamese网络的高分辨率遥感图像建筑物变化检测[J]. 中国科学院大学学报, 2023, 40(3): 380-387.
[14]	陈若男, 彭玲, 刘玉菲, 卫志超, 吕蓓茹, 陈德跃. 引入空间距离信息的城郊山区道路提取与应用[J]. 中国科学院大学学报, 2022, 39(5): 658-667.
[15]	古煜民, 阎福礼. 基于不同骨架UNet++网络的建筑物提取[J]. 中国科学院大学学报, 2022, 39(4): 512-523.