基于强化学习的免调参即插即用单光子图像重建方法

陈爽; 田烨; 付莹

doi:10.12263/DZXB.20230343

您当前的位置：

首页 >

文章列表页 >

基于强化学习的免调参即插即用单光子图像重建方法

学术论文 | 更新时间：2026-05-07

- 基于强化学习的免调参即插即用单光子图像重建方法
- Reinforcement Learning Based Tuning-free Plug-and-Play Image Reconstruction Method for Single Photon Imaging
- 电子学报 2024年52卷第10期页码：3600-3612
- 作者机构：
  
  1.北京理工大学计算机学院，北京 100081
  2.北京理工大学信息与电子学院，北京 100081
  3.北京理工大学复杂环境智能感测技术工信部重点实验室，北京 100081
- 作者简介：
  
  [ "陈爽女，1999年4月出生于北京市，硕士研究生.本科毕业于西安电子科技大学，硕士就读于北京理工大学.主要研究方向为计算机视觉.E-mail: chenshuang@bit.edu.cn" ]
  [ "田烨女，1995年6月出生于陕西省西安市，博士研究生.本科毕业于兰州大学，硕士毕业于北京大学，博士就读于北京理工大学.主要研究方向为计算成像.E-mail: 3220205110@bit.edu.cn" ]
  [ "付莹女，1987年8月出生于陕西省铜川市，北京理工大学教授，博士生导师.2009年获西安电子科技大学电子信息工程专业学士学位，2012年获清华大学自动化专业硕士学位，2015年获东京大学信息科学与技术专业博士学位.2015~2016年于东京大学从事博士后研究工作.主要研究方向为计算机视觉、多媒体图像视频分析、计算摄像学.E-mail: fuying@bit.edu.cn" ]
- 基金信息：
  
  国家自然科学基金(62331006;62171038;62088101);中央高校基本科研业务费专项资金
- DOI：10.12263/DZXB.20230343
  中图分类号： TP391;
- 收稿：2023-04-18，
  
  修回：2024-05-29，
  
  纸质出版：2024-10-25
- 稿件说明：
移动端阅览
陈爽, 田烨, 付莹. 基于强化学习的免调参即插即用单光子图像重建方法[J]. 电子学报, 2024, 52(10): 3600-3612.

CHEN Shuang, TIAN Ye, FU Ying. Reinforcement Learning Based Tuning-free Plug-and-Play Image Reconstruction Method for Single Photon Imaging[J]. Acta Electronica Sinica, 2024, 52(10): 3600-3612.
陈爽, 田烨, 付莹. 基于强化学习的免调参即插即用单光子图像重建方法[J]. 电子学报, 2024, 52(10): 3600-3612. DOI：10.12263/DZXB.20230343

CHEN Shuang, TIAN Ye, FU Ying. Reinforcement Learning Based Tuning-free Plug-and-Play Image Reconstruction Method for Single Photon Imaging[J]. Acta Electronica Sinica, 2024, 52(10): 3600-3612. DOI：10.12263/DZXB.20230343

摘要

量子图像传感器（Quanta Image Sensor，QIS）具有超高的单光子灵敏度与空间分辨率，因此其有望成为替代CMOS的下一代图像传感器.然而，从QIS获取的原始数据是二进制的图像帧，因此需要采用图像重建算法从二进制观测值中恢复原始场景.现有的方法包括基于模型的QIS图像重建和基于深度学习的QIS图像重建，其中基于模型的QIS图像重建通常需要进行大量的优化迭代，且对内部参数选择高度敏感.而基于深度学习的QIS图像重建则往往需要对每个具有细微不同的任务设计和训练单独的网络模型，灵活性欠缺并在很大程度上限制了其实用性.为了解决上述QIS图像重建方法的局限性，本文提出一种基于强化学习的免调参即插即用单光子图像重建方法.该方

法能够自适应地对不同输入图像、不同过采样倍率和重建所处的不同迭代轮次动态选取合适的参数，实现更好的重建效果.具体地，本文将即插即用框架下的QIS图像重建任务中需要手动调整的参数建模为序列决策问题，引入结合无模型强化学习和基于模型强化学习思想的算法学习一个最佳策略，以实现对于不同输入图像在每次优化迭代时的最优超参选择.在合成和真实数据集上的实验结果表明，本文提出的方法可以在过采样倍率为4、6、8三种情况下，在峰值信噪比上相较现有先进方法提高约0.44~0.60 dB，在重建的视觉效果上也能够保留更多的纹理细节.真实极暗光QIS图像数据已在

https：//github.com/ying-fu/Real-SPAD-Dataset

https://github.com/ying-fu/Real-SPAD-Dataset

公开.

Abstract

Quantum image sensor (QIS) has ultra-high single-photon sensitivity and spatial resolution

making it a promising alternative to CMOS image sensor (CIS) as the next-generation image sensor. However

image reconstruction of QIS differs from traditional image reconstruction methods

it aims to recover the original scene from binary measurements. The existing methods include model-based QIS image reconstruction and deep learning-based QIS image reconstruction. Model-based methods are largely based on optimization and are highly sensitive to the selection of hyperparameters. While deep learning-based methods require designing and training separate models for QIS image reconstruction tasks with slight variations in detail

which is inflexible and limits its usefulness to a large extent. In order to tackle the problems in QIS image reconstruction

a tuning-free plug-and-play alternating direction method of multiplier (TFPnP-ADMM) QIS image reconstruction method is proposed in this paper

which can adaptively select appropriate parameters dynamically for different input images with various oversampling factors

so as to achieve better image reconstruction performance. Specifically

in this paper

the parameters that need to be manually tuned in the QIS image reconstruction process under the plug-and-play (PnP) framework are modeled as a sequential decision problem

and a mixed model-free and model-based reinforcement learning algorithm is introduced to learn an optimal strategy

which could determine optimal hyperparameters at each itera

tion for different input images. The experimental results on synthetic dataset and real dataset demonstrate that

compared with existing state-of-the-art methods

the proposed method improves the peak signal-to-noise ratio by approximately 0.44~0.60 dB under oversampling rates of 4

and 8. Furthermore

the visual results demonstrate the superiority of the proposed method in retaining more texture details. Real extremely low light QIS image data is available at

https://github.com/ying-fu/Real-SPAD-Dataset

关键词

Keywords

references

FOSSUM E R , MA J J , MASOODIAN S , et al . The quanta image sensor: Every photon counts [J ] . Sensors , 2016 , 16 ( 8 ): 1260 .

CHAN S H . What does a one-bit quanta image sensor offer? [J ] . IEEE Transactions on Computational Imaging , 2022 , 8 : 770 - 783 .

GNANASAMBANDAM A , CHAN S H . Image classification in the dark using quanta image sensors [C ] // European Conference on Computer Vision . Cham : Springer , 2020 : 484 - 501 .

MA J J , CHAN S , FOSSUM E R . Review of quanta image sensors for ultralow-light imaging [J ] . IEEE Transactions on Electron Devices , 2022 , 69 ( 6 ): 2824 - 2839 .

LI Z P , YE J T , HUANG X , et al . Single-photon imaging over 200 km [J ] . Optica , 2021 , 8 ( 3 ): 344 - 349 .

MA S Z , GUPTA S , ULKU A C , et al . Quanta burst photography [J ] . ACM Transactions on Graphics , 2020 , 39 ( 4 ): 79 .

YANG F , LU Y M , SBAIZ L , et al . Bits from photons: Oversampled image acquisition using binary Poisson statistics [J ] . IEEE Transactions on Image Processing , 2012 , 21 ( 4 ): 1421 - 1436 .

YANG F , SBAIZ L , CHARBON E , et al . Image reconstruction in the gigavision camera [C ] // 2009 IEEE 12th International Conference on Computer Vision Workshops . Piscataway : IEEE , 2009 : 2212 - 2219 .

CHAN S H , LU Y M . Efficient image reconstruction for gigapixel quantum image sensors [C ] // 2014 IEEE Global Conference on Signal and Information Processing (GlobalSIP) . Piscataway : IEEE , 2014 : 312 - 316 .

ROJAS R A , LUO W Y , MURRAY V , et al . Learning optimal parameters for binary sensing image reconstruction algorithms [C ] // 2017 IEEE International Conference on Image Processing (ICIP) . Piscataway : IEEE , 2017 : 2791 - 2795 .

CHOI J H , ELGENDY O A , CHAN S H . Image reconstruction for quanta image sensors using deep neural networks [C ] // 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . Piscataway : IEEE , 2018 : 6543 - 6547 .

CHI Y H , GNANASAMBANDAM A , KOLTUN V , et al . Dynamic low-light imaging with quanta image sensors‍ [C ] // European Conference on Computer Vision . Cham : Springer , 2020 : 122 - 138 .

GOYAL B , GUPTA M . Photon-starved scene inference using single photon cameras [C ] // 2021 IEEE/CVF International Conference on Computer Vision (ICCV) . Piscataway : IEEE , 2021 : 2492 - 2501 .

CHAN S H , WANG X R , ELGENDY O A . Plug-and-play ADMM for image restoration: Fixed-point convergence and applications [J ] . IEEE Transactions on Computational Imaging , 2017 , 3 ( 1 ): 84 - 98 .

RYU E , LIU J L , WANG S C , et al . Plug-and-play methods provably converge with properly trained denoisers‍ [C ] // Proceedings of the 36th International Conference on Machine Learning . Cambridge : JMLR , 2019 : 5546 - 5557 .

窦诺 , 赵瑞珍 , 岑翼刚 , 等 . 基于稀疏表示的含噪图像超分辨重建方法 [J ] . 计算机研究与发展 , 2015 , 52 ( 4 ): 943 - 951 .

DOU N , ZHAO R Z , CEN Y G , et al . Noisy image super-resolution reconstruction based on sparse representation‍ [J ] . Journal of Computer Research and Development , 2015 , 52 ( 4 ): 943 - 951 . (in Chinese)

李敏 , 栗泽昊 , 张卓 , 等 . 融合局部低秩先验与Bloch流形约束的磁共振指纹重建算法 [J ] . 电子学报 , 2023 , 51 ( 2 ): 372 - 384 .

LI M , LI Z H , ZHANG Z , et al . Local low-rank and Bloch manifold regularized magnetic resonance fingerprinting reconstruction [J ] . Acta Electronica Sinica , 2023 , 51 ( 2 ): 372 - 384 . (in Chinese)

孙艳敏 , 郭强 , 张彩明 . 基于非凸低秩约束的图像修复方法 [J ] . 图学学报 , 2021 , 42 ( 3 ): 414 - 425 .

SUN Y M , GUO Q , ZHANG C M . Image inpainting using non-convex and low-rank constraint [J ] . Journal of Graphics , 2021 , 42 ( 3 ): 414 - 425 . (in Chinese)

李潇瑶 , 王炼红 , 周怡聪 , 等 . 基于四元数非局部低秩和全变分的图像混合噪声去噪算法 [J ] . 电子学报 , 2023 , 51 ( 4 ): 975 - 983 .

LI X Y , WANG L H , ZHOU Y C , et al . Image mixed denoising using quaternion-based non-local low rank and total variation [J ] . Acta Electronica Sinica , 2023 , 51 ( 4 ): 975 - 983 . (in Chinese)

OSHER S , BURGER M , GOLDFARB D , et al . An iterative regularization method for total variation-based image restoration [J ] . Multiscale Modeling & Simulation , 2005 , 4 ( 2 ): 460 - 489 .

CHEN Y , CAO W F , PANG L , et al . Hyperspectral image denoising with weighted nonlocal low-rank model and adaptive total variation regularization [J ] . IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 5544115 .

CHAN S H , ELGENDY O A , WANG X R . Images from bits: Non-iterative image reconstruction for quanta image sensors [J ] . Sensors , 2016 , 16 ( 11 ): 1961 .

GNANASAMBANDAM A , CHAN S H . HDR imaging with quanta image sensors: Theoretical limits and optimal reconstruction [J ] . IEEE Transactions on Computational Imaging , 2020 , 6 : 1571 - 1585 .

VENKATAKRISHNAN S V , BOUMAN C A , WOHLBERG B . Plug-and-play priors for model based reconstruction [C ] // 2013 IEEE Global Conference on Signal and Information Processing . Piscataway : IEEE , 2013 : 945 - 948 .

ONO S . Primal-dual plug-and-play image restoration [J ] . IEEE Signal Processing Letters , 2017 , 24 ( 8 ): 1108 - 1112 .

ZHANG K , ZUO W M , GU S H , et al . Learning deep CNN denoiser prior for image restoration [C ] // 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . Piscataway : IEEE , 2017 : 2808 - 2817 .

SUN Y , XU S Q , LI Y Z , et al . Regularized Fourier ptychography using an online plug-and-play algorithm [C ] // 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . Piscataway : IEEE , 2019 : 7665 - 7669 .

SREEHARI S , VENKATAKRISHNAN S V , WOHLBERG B , et al . Plug-and-play priors for bright field electron tomography and sparse interpolation [J ] . IEEE Transactions on Computational Imaging , 2016 , 2 ( 4 ): 408 - 423 .

ZHANG K , LI Y W , ZUO W M , et al . Plug-and-play image restoration with deep denoiser prior [J ] . IEEE Transactions on Pattern Analysis and Machine Intelligence , 2022 , 44 ( 10 ): 6360 - 6376 .

ZHANG K , ZUO W M , ZHANG L . FFDNet: Toward a fast and flexible solution for CNN based image denoising‍ [J ] . IEEE Transactions on Image Processing , 2018 , 27 ( 9 ): 4608 - 4622 .

MEINHARDT T , MOELLER M , HAZIRBAS C , et al . Learning proximal operators: Using denoising networks for regularizing inverse imaging problems [C ] // 2017 IEEE International Conference on Computer Vision (ICCV) . Piscataway : IEEE , 2017 : 1799 - 1808 .

RONNEBERGER O , FISCHER P , BROX T . U-net: Convolutional networks for biomedical image segmentation‍ [C ] // International Conference on Medical Image Computing and Computer-Assisted Intervention . Cham : Springer , 2015 : 234 - 241 .

PETERS J , SCHAAL S . Natural actor-critic [J ] . Neurocomputing , 2008 , 71 ( 7/8/9 ): 1180 - 1190 .

WANG Z Y , BAPST V , HEESS N , et al . Sample efficient actor-critic with experience replay [C ] // International Conference on Learning Representations . Toulon : OpenReview.net , 2017 : 1 - 20 .

LILLICRAP T P , HUNT J J , PRITZEL A , et al . Continuous control with deep reinforcement learning [EB/OL ] . ( 2015-09-09 )[ 2023-04-15 ] . http://arxiv.org/abs/1509.02971 http://arxiv.org/abs/1509.02971 .

BURRI S , MARUYAMA Y , MICHALET X , et al . Architecture and applications of a high resolution gated SPAD image sensor [J ] . Optics Express , 2014 , 22 ( 14 ): 17573 - 17589 .

WEI K X , AVILES-RIVERO A , LIANG J W , et al . TFPnP: Tuning-free plug-and-play proximal algorithm with applications to inverse imaging problems [J ] . The Journal of Machine Learning Research , 2022 , 23 ( 1 ): 699 - 746 .

MARTIN D , FOWLKES C , TAL D , et al . A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics [C ] // Proceedings Eighth IEEE International Conference on Computer Vision . ICCV . Piscataway : IEEE , 2001 : 416 - 423 .

KINGMA D P , BA J . Adam: A method for stochastic optimization [EB/OL ] . ( 2014-12-22 )[ 2023-04-15 ] . https://arxiv.org/abs/1412.6980 https://arxiv.org/abs/1412.6980 .

EVERINGHAM M , ESLAMI S M ALI , VAN GOOL L , et al . The pascal visual object classes challenge: A retrospective [J ] . International Journal of Computer Vision , 2015 , 111 ( 1 ): 98 - 136 .

WONG H T , LEUNG C S , HO D . Theoretical analysis and image reconstruction for multi-bit quanta image sensors [J ] . Signal Processing , 2021 , 185 : 108087 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

大语言模型增强的抗灰洞攻击海域无人机路由算法

基于图组合优化的高效社区搜索

知识数据协同的多对手智能空中博弈策略设计

基于强化学习的离散事件系统最优定向监控