[1]王志强,赵 莉,肖 锋.生成式对抗网络的图像超分辨率重建[J].西安工业大学学报,2020,(01):102-108.[doi:10.16185/j.jxatu.edu.cn.2020.01.015 ]
 WANG Zhiqiang,ZHAO Li,XIAO Feng.Image Super-Resolution Reconstruction Using a Generative Adversarial Network[J].Journal of Xi'an Technological University,2020,(01):102-108.[doi:10.16185/j.jxatu.edu.cn.2020.01.015 ]
点击复制

生成式对抗网络的图像超分辨率重建()
分享到:

《西安工业大学学报》[ISSN:1673-9965/CN:61-1458/N]

卷:
期数:
2020年01期
页码:
102-108
栏目:
信息科学与控制
出版日期:
2020-02-15

文章信息/Info

Title:
Image Super-Resolution Reconstruction Using a Generative Adversarial Network
文章编号:
1673-9965(2020)01-0102-07
作者:
王志强赵 莉肖 锋
(西安工业大学 计算机科学与工程学院,西安710021
Author(s):
WANG ZhiqiangZHAO LiXIAO Feng
(School of Computer Science and Engineering,Xi'an Technological University,Xi'an 710021,China)
关键词:
图像超分辨率 深度学习 生成式对抗网络 网络模型
Keywords:
image superresolution deep learning generative adversarial network network model
分类号:
TP312
DOI:
10.16185/j.jxatu.edu.cn.2020.01.015
文献标志码:
A
摘要:
为了提高图像生成效果,减少高频信息损失,文中提出了一种基于深度学习的生成式对抗网络模型结构,实现单幅图像超分辨率重建。文中在SRGAN方法的基础上修改了网络结构、残差网络和卷积参数,采用DIV2K数据集进行网络模型训练,利用峰值信噪比和结构相识性两种评价标准对生成的图片质量进行测试与评价。实验结果表明,相较于SRGAN方法生成的高分辨率图像,文中方法生成的图像视觉效果更好、纹理更清晰,具有更好的客观和主观评价。
Abstract:
In order to improve the effect of image generation and reduce the loss of high-frequency information,a new generative adversarial network model based on deep learning is proposed to realize super-resolution reconstruction of single image.The network structure,residual network and convolution parameters are modified based on SRGAN method.The DIV2K data set is used to train the network model.The quality of the generated images is tested and evaluated by two evaluation criteria,the peak signal-to-noise ratio(PSNR)and the structural similarity index(SSIM).Experimental results show that,compared with the SRGAN method,the images produced by the new method has better visual effect,clearer texture and better objective and subjective evaluation.

参考文献/References:

[1] GOODMAN J W.Introduction to Fourier Optics[M].New York:McGraw-Hill,1968. [2] TSAI R Y,HUANG T S.Multiple Frame Image Restoration and Registration[C]//Advances in Computer Vision and Image Processing.Greenwich:JAI Press Inc.,1984:317. [3] ALY H A,DUBOIS E.Image Up-sampling Using Total-variation Regularization with a New Observation Model[J].IEEE Transactions on Image Processing,2005,14(10):1647. [4] SU C,ZHUANG Y,HUANG L,et al.Steerable Pyramid-based Face Hallucination[J].Pattern Recognition Image Understanding for Photographs,2005,38(6):813. [5] LI X,LAM K M,QIU G,et al.Example-based Image Super-resolution With Class-specific Predictors[J].Journal of Visual Communication and Image Representation,2009,20(5):312. [6] FREEMAN W T,JONES T R,PASZTOR E C.Example-based Super-resolution[J].Computer Graphics & Applications IEEE,2002,22(2):56. [7] YANG J,WRIGHT J,HUANG T S,et al.Image Super-Resolution Via Sparse Representation[J].IEEE Transactions on Image Processing,2010,19(11):2861. [8] DONG C,LOY C C,HE K,et al.Image Super-Resolution Using Deep Convolutional Networks[J].IEEE Trans Pattern Anal Mach Intell,2014,38(2):295. [9] KIM,LEE J K,LEE K M.Accurate Image Super-resolution Using Very Deep Convolutional Networks[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:1646. [10] TAI Y,YANG J,LIU X.Image Super-Resolution Via Deep Recursive Residual Network[C]//2017 IEEE Computer Vision and Pattern Recognition.Honolulu:IEEE,2017:2790 [11] GOODFELLOW I J,POUGET-ABADIE J,MIRZA M,et al.Generative Adversarial Networks[J].Advances in Neural Information Processing Systems,2014,3:2672. [12] LEDIG C,THEIS L,HUSZAR F,et al.Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition.Honolulu:IEEE,2017:105. [13] HE K,ZHANG X,REN S,et al.Deep Residual Learning for Image Recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:770. [14] LIM B,SON S,KIM H,et al.Enhanced Deep Residual Networks for Single Image Super-Resolution[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition.Honolulu:IEEE Computer Society,2017:1132. [15] SHI W,CABALLERO J,HUSZAR F,et al.Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:1874.

相似文献/References:

[1]赵 莉,白猛猛,赵亚欣,等.生成式对抗网络的图像域转换[J].西安工业大学学报,2018,(06):645.[doi:10.16185/j.jxatu.edu.cn.2018.06.017 ]
 ZHAO Li,BAI Mengmeng,ZHAO Yaxin,et al.Image Domain Transferring by Generative Adversarial Network[J].Journal of Xi'an Technological University,2018,(01):645.[doi:10.16185/j.jxatu.edu.cn.2018.06.017 ]

备注/Memo

备注/Memo:
收稿日期:2019-05-13
基金资助:国家自然科学基金(61572392)。 第一
作者简介:王志强(1994-),男,西安工业大学硕士研究生。 通信作者简介:赵 莉(1972-),女,西安工业大学教授,主要研究方向为智能信息处理,E-mail:zhaoli1998@163.com。
(编辑、校对 肖 晨)
更新日期/Last Update: 2020-02-15