基于卡尔曼滤波的双约束CUP-VISAR压缩图像重构算法

余远平; 李海艳; 甘华权; 郑铠涛; 黄庆鑫; 理玉龙; 关赞洋; 黄运保; 景龙飞

doi:10.11884/HPLPB202335.230100

基于卡尔曼滤波的双约束CUP-VISAR压缩图像重构算法

doi: 10.11884/HPLPB202335.230100

1.
广东工业大学机电工程学院，广州 510006
2.
中国工程物理研究院激光聚变研究中心，四川绵阳 621900

基金项目: 国家自然科学基金项目(12127810， 51975125， 12105269)

详细信息

作者简介:
余远平，1762799572@qq.com

通讯作者:
李海艳，cathylhy@gdut.edu.cn

中图分类号: TP391
计量
- 文章访问数: 320
- HTML全文浏览量: 159
- PDF下载量: 60
- 被引次数: 0
出版历程
- 收稿日期: 2023-04-23
- 修回日期: 2023-06-17
- 录用日期: 2023-06-12
- 网络出版日期: 2023-06-28
- 刊出日期: 2023-08-15

Double-constrained CUP-VISAR compressed image reconstruction algorithm based on Kalman filtering

1.
School of Mechanical and Electrical Engineering, Guangdong University of Technology, Guangzhou 510006, China
2.
Laser Fusion Research Center, CAEP, Mianyang 621900, China

摘要

摘要: 针对从基于压缩超快成像（Compressed Ultrafast Photography，CUP）的任意反射面速度干涉仪（Velocity Interferometer System for Any Reflector，VISAR）中获得的压缩图像中重构出冲击波二维条纹图像的问题，提出一种基于卡尔曼滤波的双约束图像重构算法。该算法首先基于条纹图像具有的稀疏性和平滑性，将问题转化为基于小波与全变分双先验约束的优化问题，然后，考虑到实际成像的噪声问题，采用加权卡尔曼滤波对图像已有信息进行预测和调整，最后将卡尔曼滤波引入二步迭代阈值算法的迭代过程中，进而求解该双约束优化问题，实现压缩图像的精确重构。在大噪声仿真实验中，该算法重构图像的峰值信噪比和结构相似度分别提高了4.8 dB和14.81%，显著提高了图像重构质量。在实际实验中，该算法重构出了清晰的冲击波条纹图像，且将冲击波速度最大相对误差降低了9.57%和平均相对误差降低了2.2%，验证了该算法的可行性。
- 压缩超快成像 /
- 任意反射面速度干涉仪 /
- 图像重构 /
- 加权卡尔曼滤波 /
- 二步迭代阈值算法
Abstract: A dual-constrained image reconstruction algorithm based on Kalman filtering is proposed to solve the problem of reconstructing the two-dimensional shock wave fringe image from the compressed image obtained by the Velocity Interferometer System for Any Reflector (VISAR) based on Compressed Ultrafast Photography (CUP). Based on the sparsity and smoothness of fringed images, the algorithm firstly transforms the problem into an optimization problem based on wavelet and total variational double prior constraints, and then, considering the noise of actual imaging, the weighted Kalman filter is used to predict and adjust the existing information of the image, and finally the Kalman filter is introduced into the iterative process of the two-step iterative threshold algorithm, and then the double-constraint optimization problem is solved to realize the accurate reconstruction of the compressed image. In the large-noise simulation experiment, the peak signal-to-noise ratio and structural similarity of the reconstructed images of the algorithm are increased by 4.8 dB and 14.81%, respectively, which significantly improves the image reconstruction quality. In actual experiments, the algorithm reconstructs a clear shock wave fringe image and reduces the maximum relative error of shock wave velocity by 9.57% and the average relative error of shock wave velocity by 2.2%, which verifies the feasibility of the algorithm.
- compressed ultrafast photography /
- velocity interferometer system for any reflector /
- image reconstruction /
- weighted Kalman filter /
- two-step iterative threshold algorithm

HTML全文

图 1 CUP-VISAR系统

Figure 1. CUP-VISAR system

下载: 全尺寸图片幻灯片

图 2 加权卡尔曼滤波原理图

Figure 2. Weighted Kalman filter principle diagram

下载: 全尺寸图片幻灯片

图 3 去噪流程图

Figure 3. Denoising flowchart

下载: 全尺寸图片幻灯片

图 4 仿真条纹图

Figure 4. Simulated fringes

下载: 全尺寸图片幻灯片

图 5 含噪的仿真2D-VISAR条纹图

Figure 5. Simulated 2D-VISAR stripes with noise

下载: 全尺寸图片幻灯片

图 6 模拟编码图像

Figure 6. Simulated coded image

下载: 全尺寸图片幻灯片

图 7 含噪的仿真CUP-VISAR观测图像

Figure 7. Simulated CUP-VISAR observation images with noise

下载: 全尺寸图片幻灯片

图 8 不同算法的重构结果图

Figure 8. Reconstruction results of different algorithms

下载: 全尺寸图片幻灯片

图 9 重构图像的PSNR和SSIM曲线图

Figure 9. PSNR and SSIM curves of reconstructed images

下载: 全尺寸图片幻灯片

图 10 实际实验光路图

Figure 10. Diagram of actual experimental optical path

下载: 全尺寸图片幻灯片

图 11 实际实验观测数据

Figure 11. Actual experimental observation data

下载: 全尺寸图片幻灯片

图 12 不同算法的重构结果图

Figure 12. Reconstruction results of different algorithms

下载: 全尺寸图片幻灯片

图 13 从重构图像得到的线-VISAR图

Figure 13. Line-VISAR diagram obtained from the reconstructed image

下载: 全尺寸图片幻灯片

图 14 冲击波速度曲线及相对误差图

Figure 14. Shock wave velocity and relative error

下载: 全尺寸图片幻灯片

参考文献(20)

[1]	Celliers P M, Erskine D J, Sorce C M, et al. A high-resolution two-dimensional imaging velocimeter[J]. Review of Scientific Instruments, 2010, 81: 035101. doi: 10.1063/1.3310076
[2]	王峰, 关赞洋, 理玉龙, 等. 基于神光Ⅲ装置的光学诊断系统介绍[J]. 中国科学:物理学力学天文学, 2018, 48:065205 Wang Feng, Guan Zanyang, Li Yulong, et al. Optical diagnostic systems based on Shenguang Ⅲ[J]. Scientia Sinica Physica, Mechanica & Astronomica, 2018, 48: 065205
[3]	吴宇际. 激光聚变中广角冲击波速度诊断方法及相关VISAR技术研究[D]. 合肥: 中国科学技术大学, 2019 Wu Yuji. Wide-angle shock wave velocity diagnostic method and related VISAR technology in laser fusion[D]. Hefei: University of Science and Technology of China, 2019
[4]	王巧巧. 大国重器——激光惯性约束聚变[J]. 现代物理知识, 2019, 31(3):41-49 Wang Qiaoqiao. The pillars of a great power—laser inertial constraint fusion[J]. Modern Physics, 2019, 31(3): 41-49
[5]	王强强. 飞秒时间分辨条纹相机的理论和实验研究[D]. 北京: 中国科学院西安光学精密机械研究所, 2014 Wang Qiangqiang. Theoretical and experimental research on femtosecond temporal resolution streak camera[D]. Beijing: Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, 2014.
[6]	Guan Zanyang, Li Yulong, Wang Feng, et al. Study on the length of diagnostic time window of CUP-VISAR[J]. Measurement Science and Technology, 2021, 32: 125208. doi: 10.1088/1361-6501/ac29d4
[7]	Donoho D L. Compressed sensing[J]. IEEE Transactions on Information Theory, 2006, 52(4): 1289-1306. doi: 10.1109/TIT.2006.871582
[8]	Gao Liang, Liang Jinyang, Li Chiye, et al. Single-shot compressed ultrafast photography at one hundred billion frames per second[J]. Nature, 2014, 516(7529): 74-77. doi: 10.1038/nature14005
[9]	Yang Yongmei, Li Yulong, Guan Zanyang, et al. A diagnostic system toward high-resolution measurement of wavefront profile[J]. Optics Communications, 2020, 456: 124554. doi: 10.1016/j.optcom.2019.124554
[10]	Madych W R. Solutions of underdetermined systems of linear equations[R]. Shaker Heights: Institute of Mathematical Statistics, 1991: 227-238.
[11]	Qureshi M A, Deriche M. A new wavelet based efficient image compression algorithm using compressive sensing[J]. Multimedia Tools and Applications, 2016, 75(12): 6737-6754. doi: 10.1007/s11042-015-2590-9
[12]	Pandey A K, Chaudhary J, Sharma A, et al. Optimum value of scale and threshold for compression of 99mTc-MDP bone scan images using Haar wavelet transform[J]. Indian Journal of Nuclear Medicine, 2022, 37(2): 154-161. doi: 10.4103/ijnm.ijnm_170_21
[13]	查志远. 自适应范数约束图像正则化重建研究[D]. 昆明: 昆明理工大学, 2015 Zha Zhiyuan. Research on image regularization reconstruction with adaptive norm constraints[D]. Kunming: Kunming University of Science and Technology, 2015
[14]	Mahdaoui A E, Ouahabi A, Moulay M S. Image denoising using a compressive sensing approach based on regularization constraints[J]. Sensors, 2022, 22: 2199. doi: 10.3390/s22062199
[15]	Afonso M V, Bioucas-Dias J M, Figueiredo M A T. An augmented Lagrangian approach to the constrained optimization formulation of imaging inverse problems[J]. IEEE Transactions on Image Processing, 2011, 20(3): 681-695. doi: 10.1109/TIP.2010.2076294
[16]	张琦, 张慧, 潘健, 等. 一种新的卡尔曼滤波图像复原算法[J]. 湖北工业大学学报, 2022, 37(5):23-27 Zhang Qi, Zhang Hui, Pan Jian, et al. A new Kalman filter image restoration algorithm[J]. Journal of Hubei University of Technology, 2022, 37(5): 23-27
[17]	Bioucas-Dias J M, Figueiredo M A T. A new TwIST: two-step iterative shrinkage/thresholding algorithms for image restoration[J]. IEEE Transactions on Image Processing, 2007, 16(12): 2992-3004. doi: 10.1109/TIP.2007.909319
[18]	Sara U, Akter M, Uddin M S. Image quality assessment through FSIM, SSIM, MSE and PSNR—a comparative study[J]. Journal of Computer and Communications, 2019, 7(3): 8-18. doi: 10.4236/jcc.2019.73002
[19]	朱里, 李乔亮, 张婷, 等. 基于结构相似性的图像质量评价方法[J]. 光电工程, 2007, 34(11):108-113 Zhu Li, Li Qiaoliang, Zhang Ting, et al. Metric of image quality based on structural similarity[J]. Opto-Electronic Engineering, 2007, 34(11): 108-113
[20]	孙雪. 基于提升算法的9/7整数小波变换的研究及硬件实现[D]. 哈尔滨: 哈尔滨工业大学, 2013 Sun Xue. Design and implementation of 9/7 wavelet transform based lifting scheme[D]. Harbin: Harbin Institute of Technology, 2013