基于真实退化估计与高频引导的内窥镜图像超分辨率重建

李嫣; 任文琦; 张长青; 张金刚; 聂云峰

doi:10.16383/j.aas.c230070

基于真实退化估计与高频引导的内窥镜图像超分辨率重建

doi: 10.16383/j.aas.c230070

李嫣^{1, 2,},
任文琦^{3, 4,},
张长青^5,,
张金刚^3,,
聂云峰^6,

1.
中国科学院信息工程研究所北京 100093 中国
2.
中国科学院大学网络空间安全学院北京 100190 中国
3.
中国科学院大学未来技术学院北京 100039 中国
4.
中山大学网络空间安全学院深圳 518107 中国
5.
天津大学智能与计算学部天津 300350 中国
6.
布鲁塞尔自由大学应用物理与光子学系布鲁塞尔 1050 比利时

基金项目: 中国科学院网络安全和信息化专项(CAS-WX2022SF-0102), 深圳市科技计划(JCYJ20220530145209022)资助

详细信息

作者简介:
李嫣：中国科学院信息工程研究所硕士研究生. 主要研究方向为医学图像处理, 计算机视觉, 智慧医疗. E-mail: liyan1999@iie.ac.cn

任文琦：中山大学网络空间安全学院副教授. 主要研究方向为人工智能, 计算机视觉, 图像处理, 网络空间内容安全. 本文通信作者.E-mail: renwq3@mail.sysu.edu.cn

张长青：天津大学智能与计算学部副教授. 主要研究方向为机器学习, 计算机视觉, 智能医疗. E-mail: zhangchangqing@tju.edu.cn

张金刚：中国科学院大学未来技术学院副教授. 主要研究方向为智能成像技术, 医学内窥成像, 智能医学健康.E-mail: zhangjg@ucas.ac.cn

聂云峰：布鲁塞尔自由大学应用物理与光子学系教授. 主要研究方向为自由曲面光学设计算法, 成像光谱仪, 生物医学成像. E-mail: yunfeng.nie@vub.be

计量
- 文章访问数: 640
- HTML全文浏览量: 392
- PDF下载量: 167
- 被引次数: 0
出版历程
- 收稿日期: 2023-02-23
- 录用日期: 2023-10-21
- 网络出版日期: 2023-12-01
- 刊出日期: 2024-02-26

Super-resolution of Endoscopic Images Based on Real Degradation Estimation and High-frequency Guidance

LI Yan^{1, 2
,},
REN Wen-Qi^{3, 4
,},
ZHANG Chang-Qing^5
,,
ZHANG Jin-Gang^3
,,
NIE Yun-Feng^6
,

1.
Institute of Information Engineering, Chinese Academy of Sciences, Beijing 100093, China
2.
School of Cyber Security, University of Chinese Academy of Sciences, Beijing 100190, China
3.
School of Future Technology, University of Chinese Academy of Sciences, Beijing 100039, China
4.
School of Cyber Science and Technology, Sun Yat-sen University, Shenzhen 518107, China
5.
College of Intelligence and Computing, Tianjin University, Tianjin 300350, China
6.
Department of Applied Physics and Photonics, Vrije Universiteit Brussel, Brussels 1050, Belgium

Funds: Supported by Chinese Academy of Sciences Cyber Security and Informatization Project (CAS-WX2022SF-0102) and the Shenzhen Science and Technology Program (JCYJ20220530145209022)

More Information

Author Bio:
LI Yan　Master student at Institute of Information Engineering, Chinese Academy of Sciences. Her research interest covers medical image processing, computer vision, and intelligent healthcare

REN Wen-Qi　Associate professor at the School of Cyber Science and Technology, Sun Yat-sen University. His research interest covers artificial intelligence, computer vision, image processing, and content security in cyberspace. Corresponding author of this paper

ZHANG Chang-Qing　Associate professor at the College of Intelligence and Computing, Tianjin University. His research interest covers machine learning, computer vision, and intelligent healthcare

ZHANG Jin-Gang　Associate professor at the School of Future Technology, University of Chinese Academy of Sciences. His research interest covers intelligent imaging technology, medical endoscopy imaging, and intelligence medical health

NIE Yun-Feng　Professor in the Department of Applied Physics and Photonics, Vrije Universiteit Brussel. Her research interest covers freeform optical design algorithms, imaging spectrometers, and biomedical imaging

摘要

摘要: 内窥镜是诊断人体器官疾病的重要医疗设备, 然而受人体内腔环境影响, 内窥镜图像分辨率一般较低, 需对其进行超分辨处理. 目前多数基于深度学习的超分辨算法直接使用双三次插值下采样从高质量图像中获取低分辨率(Low-resolution, LR)图像以进行配对训练, 此种方式会导致纹理细节丢失, 不适用于医学图像. 为解决该问题, 针对医学内窥镜图像开发了一种新颖的退化框架, 首先从真实低质量内窥镜图像中提取丰富多样的真实模糊核与噪声模式, 之后提出一种退化注入算法, 利用提取的真实模糊核与噪声将高分辨率(High-resolution, HR)内窥镜图像退化为符合真实域的低分辨率图像. 同时, 提出一种高频引导的残差密集超分辨网络, 采用基于双频率信息交互的频率分离策略, 并设计多层级融合机制, 将提取的多级高频信息逐层嵌入残差密集模块的多层特征, 以充分恢复内窥镜图像的高频细节和低频内容. 在合成与真实数据集上的大量实验表明, 我们的方法优于对比方法, 具有更好的主客观质量评价.
- 内窥镜图像超分辨率 /
- 退化估计 /
- 高频引导 /
- 卷积神经网络
Abstract: Endoscopes are effective medical devices for diagnosing diseases of human organs. However, due to the influence of the internal cavity environment of the human body, the resolution of endoscope images is generally low. Most existing deep learning-based super-resolution algorithms directly use bicubic interpolation downsampling to obtain low-resolution (LR) images from high-quality images for paired training. However, these methods will lead to texture details loss and are not suitable for medical images. To solve this problem, this paper proposes a novel degradation framework for medical endoscopic images. First, diverse realistic blur kernels and noise patterns are extracted from real-world low-quality endoscopic images, and then a degradation injection algorithm is proposed. The extracted real blur kernels and noise degrade the high-resolution (HR) endoscopic image into a low-resolution image. In addition, this paper proposes a high-frequency guided residual dense super-resolution network, which adopts a frequency separation strategy based on dual-frequency information interaction. And a multi-level fusion mechanism is designed to embed the extracted multi-level high-frequency information into the multi-layer features of the residual dense module layer by layer. This helps recover the high-frequency details and low-frequency content of the endoscopic image. Extensive experiments on synthetic and real-world datasets show that our method outperforms the contrastive methods with better subjective and objective quality evaluations.
- Endoscopic image super-resolution /
- degradation estimation /
- high-frequency guidance /
- convolutional neural network

HTML全文

图 1 退化框架示意图

Fig. 1 Overview of the degradation framework

下载: 全尺寸图片幻灯片

图 2 HGRDN示意图

Fig. 2 Overview of the HGRDN

下载: 全尺寸图片幻灯片

图 3 常见加性高斯噪声与内窥镜噪声对比

Fig. 3 Comparison of common additive Gaussian noise with endoscopic noise

下载: 全尺寸图片幻灯片

图 4 Octave卷积之高低频率信息的交互

Fig. 4 The high-low frequency information interaction of Octave convolution

下载: 全尺寸图片幻灯片

图 5 各方法在定量测试集上的视觉结果

Fig. 5 The visual results of different methods in quantitative testsets

下载: 全尺寸图片幻灯片

图 6 各方法在定性测试集上的视觉结果

Fig. 6 The visual results of different methods in qualitative testsets

下载: 全尺寸图片幻灯片

图 7 消融实验的视觉结果

Fig. 7 The visual results of the ablation experiments

下载: 全尺寸图片幻灯片

表 1 各方法在定量测试集的客观结果

Table 1 The objective results of different methods in quantitative testsets

方法	定量测试集A		定量测试集B		定量测试集C		定量测试集D
方法	PSNR$\uparrow$	SSIM$\uparrow$	PSNR$\uparrow$	SSIM$\uparrow$	PSNR$\uparrow$	SSIM$\uparrow$	PSNR$\uparrow$	SSIM$\uparrow$
PDMSR^[55]	29.21	0.723	28.60	0.773	27.78	0.761	24.40	0.776
RealSR^[26]	28.08	0.652	28.09	0.621	25.41	0.581	25.16	0.561
RealESRGAN^[53]	31.08	0.790	30.01	0.863	32.60	0.801	32.17	0.879
USRNet^[54]	30.17	0.787	28.50	0.864	31.32	0.801	29.91	0.882
FSSR^[31]	26.46	0.670	28.31	0.663	25.93	0.612	24.32	0.574
FAWDN^[56]	31.62	0.792	32.96	0.894	32.33	0.802	33.58	0.905
BSRGAN^[30]	30.73	0.777	29.86	0.848	31.30	0.792	29.89	0.864
HGRDN (Ours)	31.78	0.797	33.22	0.902	32.61	0.808	33.90	0.913

下载: 导出CSV

表 2 各方法在定量测试集的高频结果

Table 2 The high-frequency results of different methods in quantitative testsets

方法	定量测试集A		定量测试集B
方法	高频PSNR$\uparrow$	高频SSIM$\uparrow$	高频PSNR$\uparrow$	高频SSIM$\uparrow$
PDMSR^[55]	26.95	0.573	26.07	0.573
RealSR^[26]	27.52	0.523	27.13	0.513
RealESRGAN^[53]	28.21	0.600	28.53	0.630
USRNet^[54]	27.67	0.590	27.76	0.625
FSSR^[31]	28.34	0.574	27.21	0.539
FAWDN^[56]	29.51	0.601	29.80	0.649
BSRGAN^[30]	27.13	0.543	27.65	0.580
HGRDN (Ours)	29.79	0.603	30.26	0.664

下载: 导出CSV

表 3 不同方法在定性测试集的客观结果

Table 3 The objective results of different methods in the qualitative testsets

方法	定性测试集
方法	NIQE$\downarrow$	PI$\downarrow$
PDMSR^[55]	7.96	6.73
RealSR^[26]	4.53	3.58
RealESRGAN^[53]	5.76	4.52
USRNet^[54]	9.78	8.87
FSSR^[31]	5.31	4.02
FAWDN^[56]	9.54	8.29
BSRGAN^[30]	6.74	5.57
HGRDN (Ours)	4.40	3.20

下载: 导出CSV

表 4 消融实验定量结果

Table 4 The quantitative results of the ablation experiments

方法	定量测试集A		定量测试集B
方法	PSNR$\uparrow$	SSIM$\uparrow$	PSNR$\uparrow$	SSIM$\uparrow$
去除噪声模块	30.50	0.761	32.03	0.857
去除模糊模块	30.97	0.789	32.54	0.892
去除高频模块	30.61	0.792	32.29	0.898
完整模型	31.78	0.797	33.22	0.902

下载: 导出CSV

参考文献(59)

[1]	Li B Z, Liao C R, Cai Z H, Zhou J, Zhao C, Jing L Q, et al. Femtosecond laser 3D printed micro objective lens for ultrathin fiber endoscope. Fundamental Research, DOI: 10.1016/j.fmre.2022.05.026
[2]	Yang X R, Chen Y, Tao R, Zhang Y, Liu Z W, Shi Y G. Endoscopic image deblurring and super-resolution reconstruction based on deep learning. In: Proceedings of the 2020 International Conference on Artificial Intelligence and Computer Engineering. Beijing, China: IEEE, 2020. 168−172
[3]	王知音, 张二虎, 石争浩, 段敬红. 零参考样本下的逆光图像深度学习增强方法. 中国图象图形学报, 2022, 27(5): 1589-1603 doi: 10.11834/jig.210783 Wang Zhi-Yin, Zhang Er-Hu, Shi Zheng-Hao, Duan Jing-Hong. Deep learning based backlight image enhancement method derived of zero-reference samples. Journal of Image and Graphics, 2022, 27(5): 1589-1603 doi: 10.11834/jig.210783
[4]	杨振, 邸拴虎, 赵于前, 廖苗, 曾业战. 基于级联Dense-UNet和图割的肝脏肿瘤自动分割. 电子与信息学报, 2022, 44(5): 1683-1693 Yang Zhen, Di Shuan-Hu, Zhao Yu-Qian, Liao Miao, Zeng Ye-Zhan. Automatic liver tumor segmentation based on cascaded Dense-UNet and graph cuts. Journal of Electronics & Information Technology, 2022, 44(5): 1683-1693
[5]	张芳, 赵东旭, 肖志涛, 耿磊, 吴骏, 刘彦北. 单幅图像超分辨率重建技术研究进展. 自动化学报, 2022, 48(11): 2634-2654 doi: 10.16383/j.aas.c200777 Zhang Fang, Zhao Dong-Xu, Xiao Zhi-Tao, Geng Lei, Wu Jun, Liu Yan-Bei. Research progress of single image super-resolution reconstruction technology. Acta Automatica Sinica, 2022, 48(11): 2634-2654 doi: 10.16383/j.aas.c200777
[6]	Dong C, Loy C C, He K M, Tang X O. Image super-resolution using deep convolutional networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, 38(2): 295-307 doi: 10.1109/TPAMI.2015.2439281
[7]	Ledig C, Theis L, Huszár F, Caballero J, Cunningham A, Acosta A, et al. Photo-realistic single image super-resolution using a generative adversarial network. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, USA: IEEE, 2017. 105−114
[8]	Lim B, Son S, Kim H, Nah S, Lee K M. Enhanced deep residual networks for single image super-resolution. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops. Honolulu, USA: IEEE, 2017. 1132−1140
[9]	Ahn N, Kang B, Sohn K A. Fast, accurate, and lightweight super-resolution with cascading residual network. In: Proceedings of the 15th European Conference on Computer Vision. Munich, Germany: Springer, 2018. 256−272
[10]	Qu J L, Jin K, Wang M, Huang G M. Real-time stripe noise removal method for endoscope image. In: Proceedings of the IEEE 4th International Conference on Electronics Technology. Chengdu, China: IEEE, 2021. 865−870
[11]	Chen Y H, Shi F, Christodoulou A G, Xie Y B, Zhou Z W, Li D B. Efficient and accurate MRI super-resolution using a generative adversarial network and 3D multi-level densely connected network. In: Proceedings of the 21st International Conference on Medical Image Computing and Computer-Assisted Intervention. Granada, Spain: Springer, 2018. 91−99
[12]	Park J, Hwang D, Kim K Y, Kang S K, Kim Y K, Lee J S. Computed tomography super-resolution using deep convolutional neural network. Physics in Medicine & Biology, 2018, 63(14): Article No. 145011
[13]	Mahapatra D, Bozorgtabar B. Retinal vasculature segmentation using local saliency maps and generative adversarial networks for image super resolution. arXiv preprint arXiv: 1710.04783, 2017.
[14]	蒋希, 袁奕萱, 王雅萍, 肖振祥, 朱美芦, 陈泽华, 等. 中国医学影像人工智能20年回顾和展望. 中国图象图形学报, 2022, 27(3): 655-671 Jiang Xi, Yuan Yi-Xuan, Wang Ya-Ping, Xiao Zhen-Xiang, Zhu Mei-Lu, Chen Ze-Hua, et al. A 20-year retrospect and prospect of medical imaging artificial intelligence in China. Journal of Image and Graphics, 2022, 27(3): 655-671
[15]	Gu X G, Zhou F X, Chen R F, Ren X Z, Zhou W J. Endoscopic single image super-resolution based on transformer and convolutional neural network. In: Proceedings of the 7th International Conference on Life System Modeling and Simulation. Hangzhou, China: Springer, 2021. 24−32
[16]	Song X W, Tang H, Yang C F, Zhou G Q, Wang Y G, Huang X J, et al. Deformable transformer for endoscopic video super-resolution. Biomedical Signal Processing and Control, 2022, 77: Article No. 103827
[17]	Turan M. A generative adversarial network based super-resolution approach for capsule endoscopy images. Medicine Science, 2021, 10(3): 1002-1007 doi: 10.5455/medscience.2021.06.218
[18]	Almalioglu Y, Ozyoruk K B, Gokce A, Incetan K, Gokceler G I, Simsek M A, et al. EndoL2H: Deep super-resolution for capsule endoscopy. IEEE Transactions on Medical Imaging, 2020, 39(12): 4297-4309 doi: 10.1109/TMI.2020.3016744
[19]	Dai T, Cai J R, Zhang Y B, Xia S T, Zhang L. Second-order attention network for single image super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, USA: IEEE, 2019. 11057−11066
[20]	Pan J S, Liu S F, Sun D Q, Zhang J W, Liu Y, Ren J, et al. Learning dual convolutional neural networks for low-level vision. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE, 2018. 3070−3079
[21]	He X Y, Mo Z T, Wang P S, Liu Y, Yang M Y, Cheng J. ODE-inspired network design for single image super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, USA: IEEE, 2019. 1732−1741
[22]	Hu X C, Mu H Y, Zhang X Y, Wang Z L, Tan T N, Sun J. Meta-SR: A magnification-arbitrary network for super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, USA: IEEE, 2019. 1575−1584
[23]	Li Z, Yang J L, Liu Z, Yang X M, Jeon G, Wu W. Feedback network for image super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, USA: IEEE, 2019. 3862−3871
[24]	Qiu Y J, Wang R X, Tao D P, Cheng J. Embedded block residual network: A recursive restoration model for single-image super-resolution. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. Seoul, South Korea: IEEE, 2019. 4179−4188
[25]	Yin X, Tai Y, Huang Y G, Liu X M. FAN: Feature adaptation network for surveillance face recognition and normalization. In: Proceedings of the 15th Asian Conference on Computer Vision. Kyoto, Japan: Springer, 2020. 301−319
[26]	Ji X Z, Cao Y, Tai Y, Wang C J, Li J L, Huang F Y. Real-world super-resolution via kernel estimation and noise injection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Seattle, USA: IEEE, 2020. 1914−1923
[27]	Wang W, Zhang H C, Yuan Z H, Wang C H. Unsupervised real-world super-resolution: A domain adaptation perspective. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. Montreal, Canada: IEEE, 2021. 4298−4307
[28]	Zhang K, Zuo W M, Zhang L. Learning a single convolutional super-resolution network for multiple degradations. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE, 2018. 3262−3271
[29]	Zhang K, Zuo W M, Zhang L. Deep plug-and-play super-resolution for arbitrary blur kernels. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, USA: IEEE, 2019. 1671−1681
[30]	Zhang K, Liang J Y, Van Gool L, Timofte R. Designing a practical degradation model for deep blind image super-resolution. In: Proceedings of IEEE/CVF International Conference on Computer Vision. Montreal, Canada: IEEE, 2021. 4771−4780
[31]	Fritsche M, Gu S H, Timofte R. Frequency separation for real-world super-resolution. In: Proceedings of IEEE/CVF International Conference on Computer Vision Workshop. Seoul, South Korea: IEEE, 2019. 3599−3608
[32]	Lugmayr A, Danelljan M, Timofte R. Unsupervised learning for real-world super-resolution. In: Proceedings of IEEE/CVF International Conference on Computer Vision Workshop. Seoul, South Korea: IEEE, 2019. 3408−3416
[33]	Chen S J, Han Z, Dai E Y, Jia X, Liu Z L, Liu X, et al. Unsupervised image super-resolution with an indirect supervised path. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Seattle, USA: IEEE, 2020. 1924−1933
[34]	Xie W B, Song D H, Xu C, Xu C J, Zhang H, Wang Y H. Learning frequency-aware dynamic network for efficient super-resolution. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. Montreal, Canada: IEEE, 2021. 4288−4297
[35]	Yun J S, Yoo S B. Single image super-resolution with arbitrary magnification based on high-frequency attention network. Mathematics, 2022, 10(2): Article No. 275 doi: 10.3390/math10020275
[36]	Xu R Y, Kang X J, Li C X, Chen H, Ming A L. DCT-FANet: DCT based frequency attention network for single image super-resolution. Displays, 2022, 74: Article No. 102220 doi: 10.1016/j.displa.2022.102220
[37]	Guo T T, Mousavi H S, Monga V. Adaptive transform domain image super-resolution via orthogonally regularized deep networks. IEEE Transactions on Image Processing, 2019, 28(9): 4685-4700 doi: 10.1109/TIP.2019.2913500
[38]	Hung K W, Wang K, Jiang J M. Image up-sampling using deep cascaded neural networks in dual domains for images down-sampled in DCT domain. Journal of Visual Communication and Image Representation, 2018, 56: 144-149 doi: 10.1016/j.jvcir.2018.09.005
[39]	Pang Y X, Li X, Jin X, Wu Y J, Liu J Z, Liu S, et al. FAN: Frequency aggregation network for real image super-resolution. In: Proceedings of the European Conference on Computer Vision. Glasgow, UK: Springer, 2020. 468−483
[40]	Liu Z B, Yuan L, Sun L. Frequency separation-based multi-scale cascading residual block network for image super resolution. Multimedia Tools and Applications, 2022, 81(5): 6827-6848 doi: 10.1007/s11042-021-11724-z
[41]	Chen Y P, Fan H Q, Xu B, Yan Z C, Kalantidis Y, Rohrbach M, et al. Drop an octave: Reducing spatial redundancy in convolutional neural networks with octave convolution. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. Seoul, South Korea: IEEE, 2019. 3434−3443
[42]	Bell-Kligler S, Shocher A, Irani M. Blind super-resolution kernel estimation using an internal-GAN. In: Proceedings of the 33rd International Conference on Neural Information Processing Systems. Vancouver, Canada: Curran Associates Inc., 2019. Article No. 26
[43]	Chen J W, Chen J W, Chao H Y, Yang M. Image blind denoising with generative adversarial network based noise modeling. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE, 2018. 3155−3164
[44]	Sudiro S A, Kardian A R, Madenda S, Hermanto L. Mean and variance statistic for image processing on FPGA. International Journal of Applied Science and Engineering, 2021, 18(1): Article No. 2020115
[45]	陈晔曜, 蒋刚毅, 邵华, 姜浩, 郁梅. 高动态范围图像融合过程中的噪声抑制算法. 光电工程, 2018, 45(7): Article No. 180083 Chen Ye-Yao, Jiang Gang-Yi, Shao Hua, Jiang Hao, Yu Mei. Noise suppression algorithm in the process of high dynamic range image fusion. Opto-Electronic Engineering, 2018, 45(7): Article No. 180083
[46]	Luo Z X, Huang Y, Li S, Wang L, Tan T N. Unfolding the alternating optimization for blind super resolution. In: Proceedings of the 34th International Conference on Neural Information Processing Systems. Vancouver, Canada: Curran Associates Inc., 2020. Article No. 473
[47]	Wang X T, Yu K, Wu S X, Gu J J, Liu Y H, Dong C, et al. ESRGAN: Enhanced super-resolution generative adversarial networks. In: Proceedings of the European Conference on Computer Vision. Munich, Germany: Springer, 2018. 63−79
[48]	Zhang Y L, Tian Y P, Kong Y, Zhong B N, Fu Y. Residual dense network for image super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE, 2018. 2472−2481
[49]	Zhang Y L, Li K P, Li K, Wang L C, Zhong B N, Fu Y. Image super-resolution using very deep residual channel attention networks. In: Proceedings of the 15th European Conference on Computer Vision. Munich, Germany: Springer, 2018. 294−310
[50]	Kingma D P, Ba J. Adam: A method for stochastic optimization. In: Proceedings of the 3rd International Conference on Learning Representations. San Diego, USA: ICLR, 2015.
[51]	Pogorelov K, Randel K R, Griwodz C, Eskeland S L, de Lange T, Johansen D, et al. KVASIR: A multi-class image dataset for computer aided gastrointestinal disease detection. In: Proceedings of the 8th ACM on Multimedia Systems Conference. Taipei, China: Association for Computing Machinery, 2017. 164−169
[52]	Smedsrud P H, Thambawita V, Hicks S A, Gjestang H, Nedrejord O O, Naess E, et al. Kvasir-Capsule, a video capsule endoscopy dataset. Scientific Data, 2021, 8(1): Article No. 142 doi: 10.1038/s41597-021-00920-z
[53]	Wang X T, Xie L B, Dong C, Shan Y. Real-ESRGAN: Training real-world blind super-resolution with pure synthetic data. In: Proceedings of IEEE/CVF International Conference on Computer Vision Workshops. Montreal, Canada: IEEE, 2021. 1905−1914
[54]	Zhang K, Van Gool L, Timofte R. Deep unfolding network for image super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle, USA: IEEE, 2020. 3214−3223
[55]	Luo Z X, Huang Y, Li S, Wang L, Tan T N. Learning the degradation distribution for blind image super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans, USA: IEEE, 2022. 6063−6072
[56]	Chen L H, Yang X M, Jeon G, Anisetti M, Liu K. A trusted medical image super-resolution method based on feedback adaptive weighted dense network. Artificial Intelligence in Medicine, 2020, 106: Article No. 101857 doi: 10.1016/j.artmed.2020.101857
[57]	Gonzalez R C, Woods R E [著], 阮秋琦, 阮宇智[译]. 数字图像处理. 第3版. 北京: 电子工业出版社, 2011. 169−182 Gonzalez R C, Woods R E [Author], Ruan Qiu-Qi, Ruan Yu-Zhi [Translator]. Digital Image Processing. Beijing: Publishing House of Electronics Industry, 2011. 169−182
[58]	Mittal A, Soundararajan R, Bovik A C. Making a “completely blind” image quality analyzer. IEEE Signal Processing Letters, 2013, 20(3): 209-212 doi: 10.1109/LSP.2012.2227726
[59]	Blau Y, Mechrez R, Timofte R, Michaeli T, Zelnik-Manor L. The 2018 PIRM challenge on perceptual image super-resolution. In: Proceedings of the European Conference on Computer Vision. Munich, Germany: Springer, 2018. 334−355