[1]邱德府,郑力新,谢炜芳,等.深度学习下的高效单幅图像超分辨率重建方法[J].华侨大学学报(自然科学版),2019,40(5):668-673.[doi:10.11830/ISSN.1000-5013.201905029]
　QIU Defu,ZHENG Lixin,XIE Weifang,et al.Efficient Single Image Super-Resolution Reconstruction Method Under Deep Learning[J].Journal of Huaqiao University(Natural Science),2019,40(5):668-673.[doi:10.11830/ISSN.1000-5013.201905029]
点击复制

深度学习下的高效单幅图像超分辨率重建方法()
分享到：

《华侨大学学报（自然科学版）》[ISSN:1000-5013/CN:35-1079/N]

卷:
第40卷
期数:
2019年第5期
页码:
668-673
栏目:
出版日期:
2019-09-20

文章信息/Info

Title:
Efficient Single Image Super-Resolution Reconstruction Method Under Deep Learning
文章编号:
1000-5013(2019)05-0668-06
作者:
邱德府1; 2;  郑力新1; 2;  谢炜芳1; 2;  朱建清1; 2
1. 华侨大学 工学院, 福建 泉州 362021;2. 华侨大学 工业智能化技术与系统福建省高校工程研究中心, 福建 泉州 362021
Author(s):
QIU Defu1; 2;  ZHENG Lixin1; 2;  XIE Weifang1; 2;  ZHU Jianqing1; 2
1. College of Engineering, Huaqiao University, Quanzhou 362021, China; 2. Industrial Intelligence and System Fujian University Engineering Research Center, Huaqiao University, Quanzhou 362021, China
关键词:
深度学习;  超分辨率重建;  卷积神经网络;  亚像素卷积;  风格转移
Keywords:
deep learning;  super-resolution reconstruction;  convolutional neural networks;  sub-pixel convolution;  style transfer
分类号:
TP391
DOI:
10.11830/ISSN.1000-5013.201905029
文献标志码:
A
摘要:
提出一种基于深度学习的高效单幅图像超分辨率重建方法,增加一个亚像素卷积层和一个替换的级联卷积,即设计一个具有合适深度的卷积神经网络,以保证图像的重建质量,并采用级联小卷积核提高运行速度.在标准的公共数据集上进行实验测试,结果表明:与亚像素卷积神经网络(ESPCN)算法相比,所提算法重建的高分辨率图像的质量和速度皆显著提高;将其应用于实际项目中,可端到端地重建低分辨率服装风格图像,获得高分辨率图像.
Abstract:
An efficient single-image super-resolution reconstruction method based on deep learning is proposed. A sub-pixel convolution layer and a replacement concatenated convolution are added to design a convolutional neural network with appropriate depth to ensure image reconstruction quality. And use cascaded small convolution kernel to enhance the running speed. The results show that the quality and speed of high-resolution images reconstructed by the proposed algorithm are significantly improved compared to sub-pixel convolutional neural networks(ESPCN)algorithm. Using the proposed method in practical projects, we can achieve end-to-end reconstruction of low-resolution clothing style images, and obtain high-resolution images.

参考文献/References:

[1] 邢晓羊.基于深度学习的医学图像超分辨率重建[D].成都:成都信息工程大学,2018.
[2] ZHOU Fei,YANG Wenming,LIAO Qingmin.Interpolation-based image super-resolution using multisurface fifitting[J].IEEE Transactions on Image Processing,2012,21(7):3312-3318.DOI:10.1109/TIP.2012.2189576.
[3] LIN Zhouchen,SHUM H Y.Fundamental limits of reconstruction based super resolution algorithms under local translation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2004,26(1):83-97.DOI:10.1109/tpami.2004.1261081.
[4] 潘宗序,禹晶,胡少兴,等.基于多尺度结构自相似性的单幅图像超分辨率算法[J].自动化学报,2014,40(4):594-603.DOI:10.3724/SP.J.1004.2014.00594.
[5] 席志红,侯彩燕,袁昆鹏,等.基于深层残差网络的加速图像超分辨率重建[J].光学学报,2019,39(2):89-98.
[6] 刘哲,黄文准,乌伟.基于级联线性回归的快速单幅图像超分辨率技术[J].红外技术,2018,40(9):894-901.
[7] 曾凯,丁世飞.图像超分辨率重建的研究进展[J].计算机工程与应用,2017,53(16):29-35.DOI:10.3778/j.issn.1002-8331.1705-0097.
[8] ZHANG Yongqin,LIU Jianying,YANG Wenhan,et al.Image super-resolution based on structure-modulated sparse representation[J].IEEE Transactions on Image Processing,2015,24(9):2797-2810.DOI:10.1109/TIP.2015.2431435.
[9] YANG C Y,HUANG Jiabin,YANG M H.Exploiting self-similarities for single frame super-resolution[C]//Asian Conference on Computer Vision.Queenstown:[s.n.],2010:497-510.DOI:10.1007/978-3-642-19318-7_39.
[10] YANG Jianchao,WRIGHT J,HUANG T S,et al.Image super-resolution via sparse representation[J].IEEE Transactions on Image Processing,2010,19(11):2861-2873.DOI:10.1109/TIP.2010.2050625.
[11] TIMOFTE R,SMET V D,GOOL L V.Anchored neighborhood regression for fast example-based super-resolution[C]//IEEE International Conference on Computer Vision.Sydney:IEEE Press,2013:1920-1927.DOI:10.1109/iccv.2013.241.
[12] DONG Chao,CHEN C L,HE Kaiming,et al.Image super-resolution using deep convolutional networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(2):295-307.DOI:10.1109TPAMI.2015.2439281.
[13] SHI Wenzhe,CABALLERO J,HUSZAR F,et al.Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network[C]//IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Press,2016:1874-1883.DOI:10.1109/CVPR.2016.207.
[14] KIM J,LEE J K,LEE K M.Accurate image super-resolution using very deep convolutional networks[C]//IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Press,2016:1646-1654.DOI:10.1109/CVPR.2016.182.
[15] 李云红,王珍,张凯兵,等.基于学习的图像超分辨重建方法综述[J].计算机工程与应用,2018,54(15):13-21.DOI:10.3778/j.issn.1002-8331.1805-0038.
[16] 张圣祥,郑力新,朱建清,等.采用深度学习的快速超分辨率图像重建方法[J].华侨大学学报(自然科学版),2019,40(2):245-250.DOI:10.11830/ISSN.1000-5013.201804064.
[17] 刘月峰,杨涵晰,蔡爽,等.基于改进卷积神经网络的单幅图像超分辨率重建方法[J].计算机应用,2019,39(5):1440-1447.DOI:10.11772/j.issn.1001-9081.2018091887.

备注/Memo

备注/Memo:
收稿日期: 2019-05-16
通信作者: 郑力新(1967-),男,教授,博士,主要从事运动控制、机器视觉、图像处理与模式识别的研究.E-mail:zlx@hqu.edu.cn.
基金项目: 福建省泉州市高层次人才创新创业项目(2017G036); 国家自然科学基金青年基金资助项目(61602191); 福建省厦门市科技计划项目(3502Z20173045); 华侨大学研究生科研创新能力培育计划资助项目(17014084010)
更新日期/Last Update: 2019-09-20