Unsupervised deep learning method for single image super-resolution of the thick pinhole imaging system using deep image prior
Thick pinhole imaging system is widely used for diagnosing intense pulsed radiation sources. However, owing to the trade-off among spatial resolution, field of view (FOV) and signal-to-noise ratio (SNR), the imaging system normally falls short in achieving high-precision spatial diagnosis. In this p...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Nuclear Engineering and Technology |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1738573324003863 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841545587572342784 |
|---|---|
| author | Guoguang Li Liang Sheng Baojun Duan Yang Li Dongwei Hei Qingzi Xing |
| author_facet | Guoguang Li Liang Sheng Baojun Duan Yang Li Dongwei Hei Qingzi Xing |
| author_sort | Guoguang Li |
| collection | DOAJ |
| description | Thick pinhole imaging system is widely used for diagnosing intense pulsed radiation sources. However, owing to the trade-off among spatial resolution, field of view (FOV) and signal-to-noise ratio (SNR), the imaging system normally falls short in achieving high-precision spatial diagnosis. In this paper, we propose an unsupervised deep learning method for single image super-resolution (SISR) of the thick pinhole imaging system. The point spread function (PSF) of the imaging system is obtained by analytical calculation and Monte Carlo simulation methods, and the mathematical model of the imaging system is established using a linear equation. To solve the ill-posed inverse problem, we adopt randomly initialized deep convolutional neural networks (DCNNs) as an image prior without pre-training, which is named deep image prior (DIP). The results demonstrate that, by utilizing the SISR technique to increase the number of pixels in reconstructed images, the proposed DIP algorithm can mitigate the spatial resolution degradation caused by an insufficient spatial sampling frequency of the camera. Compared with various classical algorithms, the proposed DIP algorithm exhibits superior capabilities in recovering high-frequency signals and suppressing ringing artifacts. Furthermore, the convergence and robustness of the proposed DIP algorithm under different random seeds and SNR conditions are also verified. |
| format | Article |
| id | doaj-art-6e886ce9c9f14c289937fd249723cbe4 |
| institution | Kabale University |
| issn | 1738-5733 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Nuclear Engineering and Technology |
| spelling | doaj-art-6e886ce9c9f14c289937fd249723cbe42025-01-12T05:24:36ZengElsevierNuclear Engineering and Technology1738-57332025-01-01571103139Unsupervised deep learning method for single image super-resolution of the thick pinhole imaging system using deep image priorGuoguang Li0Liang Sheng1Baojun Duan2Yang Li3Dongwei Hei4Qingzi Xing5Key Laboratory of Particle & Radiation Imaging (Tsinghua University), Ministry of Education, Beijing, 100084, China; Laboratory for Advanced Radiation Sources and Application, Tsinghua University, Beijing, 100084, China; Department of Engineering Physics, Tsinghua University, Beijing, 100084, ChinaState Key Laboratory of Intense Pulsed Radiation Simulation and Effect, Northwest Institute of Nuclear Technology, Xi'an, 710024, China; Corresponding author. Northwest Institute of Nuclear Technology, No. 28 Pingyu Road, Baqiao District, Xi'an, 710024, China.State Key Laboratory of Intense Pulsed Radiation Simulation and Effect, Northwest Institute of Nuclear Technology, Xi'an, 710024, ChinaState Key Laboratory of Intense Pulsed Radiation Simulation and Effect, Northwest Institute of Nuclear Technology, Xi'an, 710024, ChinaState Key Laboratory of Intense Pulsed Radiation Simulation and Effect, Northwest Institute of Nuclear Technology, Xi'an, 710024, ChinaKey Laboratory of Particle & Radiation Imaging (Tsinghua University), Ministry of Education, Beijing, 100084, China; Laboratory for Advanced Radiation Sources and Application, Tsinghua University, Beijing, 100084, China; Department of Engineering Physics, Tsinghua University, Beijing, 100084, ChinaThick pinhole imaging system is widely used for diagnosing intense pulsed radiation sources. However, owing to the trade-off among spatial resolution, field of view (FOV) and signal-to-noise ratio (SNR), the imaging system normally falls short in achieving high-precision spatial diagnosis. In this paper, we propose an unsupervised deep learning method for single image super-resolution (SISR) of the thick pinhole imaging system. The point spread function (PSF) of the imaging system is obtained by analytical calculation and Monte Carlo simulation methods, and the mathematical model of the imaging system is established using a linear equation. To solve the ill-posed inverse problem, we adopt randomly initialized deep convolutional neural networks (DCNNs) as an image prior without pre-training, which is named deep image prior (DIP). The results demonstrate that, by utilizing the SISR technique to increase the number of pixels in reconstructed images, the proposed DIP algorithm can mitigate the spatial resolution degradation caused by an insufficient spatial sampling frequency of the camera. Compared with various classical algorithms, the proposed DIP algorithm exhibits superior capabilities in recovering high-frequency signals and suppressing ringing artifacts. Furthermore, the convergence and robustness of the proposed DIP algorithm under different random seeds and SNR conditions are also verified.http://www.sciencedirect.com/science/article/pii/S1738573324003863Thick pinhole imaging systemPoint spread functionSingle image super-resolutionUnsupervised deep learningDeep image prior |
| spellingShingle | Guoguang Li Liang Sheng Baojun Duan Yang Li Dongwei Hei Qingzi Xing Unsupervised deep learning method for single image super-resolution of the thick pinhole imaging system using deep image prior Nuclear Engineering and Technology Thick pinhole imaging system Point spread function Single image super-resolution Unsupervised deep learning Deep image prior |
| title | Unsupervised deep learning method for single image super-resolution of the thick pinhole imaging system using deep image prior |
| title_full | Unsupervised deep learning method for single image super-resolution of the thick pinhole imaging system using deep image prior |
| title_fullStr | Unsupervised deep learning method for single image super-resolution of the thick pinhole imaging system using deep image prior |
| title_full_unstemmed | Unsupervised deep learning method for single image super-resolution of the thick pinhole imaging system using deep image prior |
| title_short | Unsupervised deep learning method for single image super-resolution of the thick pinhole imaging system using deep image prior |
| title_sort | unsupervised deep learning method for single image super resolution of the thick pinhole imaging system using deep image prior |
| topic | Thick pinhole imaging system Point spread function Single image super-resolution Unsupervised deep learning Deep image prior |
| url | http://www.sciencedirect.com/science/article/pii/S1738573324003863 |
| work_keys_str_mv | AT guoguangli unsuperviseddeeplearningmethodforsingleimagesuperresolutionofthethickpinholeimagingsystemusingdeepimageprior AT liangsheng unsuperviseddeeplearningmethodforsingleimagesuperresolutionofthethickpinholeimagingsystemusingdeepimageprior AT baojunduan unsuperviseddeeplearningmethodforsingleimagesuperresolutionofthethickpinholeimagingsystemusingdeepimageprior AT yangli unsuperviseddeeplearningmethodforsingleimagesuperresolutionofthethickpinholeimagingsystemusingdeepimageprior AT dongweihei unsuperviseddeeplearningmethodforsingleimagesuperresolutionofthethickpinholeimagingsystemusingdeepimageprior AT qingzixing unsuperviseddeeplearningmethodforsingleimagesuperresolutionofthethickpinholeimagingsystemusingdeepimageprior |