Intensity and phase imaging through scattering media via deep despeckle complex neural networks

Liu, Shuai; Sha, Hao; Dong, Jiuyang; Huang, Yue; Zhao, Yong; Yao, Xu‐Ri; Qin, Ping; Li, Xiu; Lin, Xing; Zhang, Yongbing

doi:10.1016/j.optlaseng.2022.107196

Cited by 5 publications

(3 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep learning executes advanced inference tasks using computers and presents great achievements in imaging through scattering the media. 80 The optical diffractive deep neural networks built based on passive structures perform complex functions using computer-based neural networks. 68 The scattering in optical diffusers leads to the degeneration of light intensity and phase; to recover the intensity and phase information, deep neural networks have been employed.…”

Section: Theory and Measurementsmentioning

confidence: 99%

A comprehensive review of optical diffusers: progress and prospects

2023

View full text Add to dashboard Cite

show abstract

Section: Theory and Measurementsmentioning

confidence: 99%

A comprehensive review of optical diffusers: progress and prospects

2023

View full text Add to dashboard Cite

show abstract

“…Particularly for objects that are transparent, reconstructing the phase distribution can lead to visualization of the spatial distribution such as thickness and refractive index. Researchers have also developed the reconstruction of phase information through scattering media [1][2][3][4][5].…”

Section: Introductionmentioning

confidence: 99%

Imaging phase objects through diffusers based on lensless digital holography

Watanabe,

Asada,

Kiyosumi

et al. 2024

Opt Rev

View full text Add to dashboard Cite

Imaging of phase objects behind scattering media is a challenging task. Intensity imaging through diffusers can be achieved based on digital holography by obtaining the complex amplitude of the diffuser in advance. As described in this paper, we experimentally demonstrate the reconstructed images of phase objects behind diffusers with different diffusion angles by digital holography. Using the complex amplitude information of the diffuser to correct the complex amplitude information of the object through the diffuser, the phase distribution of the object is obtainable behind the diffuser. Imaging of phase objects behind diffusers has been verified through experiments using a plano-convex lens and a wedge substrate as phase objects with various scattering angles. Quantitative analyses of the phase objects are performed. The lens shape can be visualized from the known refractive index. Moreover, the curvature radius can be estimated.

show abstract

“…13 Current deep learning models, however, focus on restoring simple geometric patterns and "one-to-one" mapping datasets. To overcome MMF's variability and randomness in continuous transmission and imaging, scalable semi-supervised learning model, deep de-speckle complex neural network, convolutional neural network (CNN), and multi-scale memory dynamic-learning network have been designed by Fan, 14 Liu, 15 Xu, 16 and Li et al 17 While ARTICLE pubs.aip.org/aip/adv these networks have achieved transient and short-term continuous transmission and imaging for complex images through MMFs, a more robust and long-term MMF de-speckle neural network is still needed for practical scenarios, such as long-term visualization and minimally invasive surgery. This paper presents a multi-channel symmetric network (MCSNet) architecture, comprising U-Net 18 and ConvNeXt Block, 19 to predict the effective features of speckles.…”

mentioning

confidence: 99%

Learning to see high-density random images long-term transmitted in multimode fiber

Li,

Song,

et al. 2024

AIP Advances

View full text Add to dashboard Cite

An improved multi-channel symmetric network (MCSNet) is proposed to reconstruct high-channel-density random images after long-term transmission through multimode fibers (MMFs). Temporal correlation within a period of 25 minutes is calculated to investigate the time-varying characteristics of speckles. The results demonstrated that due to noise accumulation along the MMF path, the quality of speckles deteriorates significantly after long-term transmission. The MCSNet integrates U-Net and ConvNeXt Block, which enables to more fully extract the features of each channel within the entire speckle. After being trained by different random image datasets within the initial moment, tests on random images and realistic scenes of endoscopic surgery after 25 min of transmission are carried out, and all of them demonstrate a near-perfect reconstruction performance and superior scalability, which indicates that MCSNet is suitable for long-term imaging demodulation of endoscopes.

show abstract

Intensity and phase imaging through scattering media via deep despeckle complex neural networks

Cited by 5 publications

References 59 publications

A comprehensive review of optical diffusers: progress and prospects

A comprehensive review of optical diffusers: progress and prospects

Imaging phase objects through diffusers based on lensless digital holography

Learning to see high-density random images long-term transmitted in multimode fiber

Contact Info

Product

Resources

About