###
Journal of Software:2018.29(4):926-934

基于深度反卷积神经网络的图像超分辨率算法
彭亚丽,张鲁,张钰,刘侍刚,郭敏
(现代教学技术教育部重点实验室(陕西师范大学), 陕西 西安 710062;陕西省教学信息技术工程实验室(陕西师范大学), 陕西 西安 710119;陕西师范大学 计算机科学学院, 陕西 西安 710119)
Deep Deconvolution Neural Network for Image Super-Resolution
PENG Ya-Li,ZHANG Lu,ZHANG Yu,LIU Shi-Gang,GUO Min
(Key Laboratory of Modern Teaching Technology, Ministry of Education(Shaanxi Normal University), Xi'an 710062, China;Engineering Laboratory of Teaching Information Technology of Shaanxi Province(Shaanxi Normal University), Xi'an 710119, China;School of Computer Science, Shaanxi Normal University, Xi'an 710119, China)
Abstract
Chart / table
Reference
Similar Articles
Article :Browse 1977   Download 3606
Received:April 29, 2017    Revised:June 26, 2017
> 中文摘要: 图像超分辨率一直是底层视觉领域的研究热点,现有基于卷积神经网络的方法直接利用传统网络模型,未对图像超分辨率属于回归问题这一本质进行优化,其网络学习能力较弱,训练时间较长,重建图像的质量仍有提升空间.针对这些问题,提出了基于深度反卷积神经网络的图像超分辨率算法,该算法利用反卷积层对低分辨率图像进行上采样处理,再经深度映射消除由反卷积层造成的噪声和伪影现象,使用残差学习降低网络复杂度,同时避免了因网络过深导致的网络退化问题.在Set 5、Set 14等测试集中,所提算法的PSNR、SSIM、IFC这3项评价指标都优于FSRCNN,重建图像的视觉效果同样验证了该算法出色的性能.
Abstract:Image super resolution is a research hot spot in the field of low level vision. The existing methods based on convolutional neural network do not optimize the image super resolution as a regression problem. These methods are weak in learning ability and require too much time in training step, also leaving room for improvement in the quality of image reconstruction. To solve above mentioned problems, this article proposes a method based on deep deconvolution neural network, which first upsamples low resolution image by deconvolution layer, and then uses deep mapping to eliminate the noise and artifacts caused by deconvolution layer. The residual learning reduces the network complexity and avoids the network degradation caused by the depth network. In Set 5, Set 14 and other datasets, the presented method performs better than FSRCNN in PSNR, SSIM, IFC and visual.
文章编号:     中图分类号:    文献标志码:
基金项目:国家自然科学基金(61672333,61741208,61402274,61772325);陕西省工业科技攻关项目(2016GY-081);教育部高等教育司产学合作协同育人项目(201701023062);陕西省自然科学基金(2017JQ6074);陕西省农业攻关项目(2016NY-176);陕西省重点科技创新团队计划(2014KTC-18);教育部科技发展中心"云数融合科教创新"基金(2017A07053);陕西师范大学学习科学交叉学科培育计划;中央高校基本科研业务费专项资金(2017CSY024,GK201603091,GK201703054) 国家自然科学基金(61672333,61741208,61402274,61772325);陕西省工业科技攻关项目(2016GY-081);教育部高等教育司产学合作协同育人项目(201701023062);陕西省自然科学基金(2017JQ6074);陕西省农业攻关项目(2016NY-176);陕西省重点科技创新团队计划(2014KTC-18);教育部科技发展中心"云数融合科教创新"基金(2017A07053);陕西师范大学学习科学交叉学科培育计划;中央高校基本科研业务费专项资金(2017CSY024,GK201603091,GK201703054)
Foundation items:National Natural Science Foundation of China (61672333, 61741208, 61402274, 61772325); Key Science and Technology Program of Shaanxi Province of China (2016GY-081); Industry University Cooperative Education Project of Higher Education Department of the Ministry of Education (201701023062); Natural Science Foundation of Shaanxi Province, China (2017JQ6074); Science Research and Development Program of Shaanxi Province of China (2016NY-176); Program of Key Science and Technology Innovation Team in Shaanxi Province (2014KTC-18); Fund for Integration of Cloud Computing and Big Data of Science and Technology Development Center of the Ministry of Education (2017A07053); Interdisciplinary Incubation Project of Learning Science of Shaanxi Normal University; Fundamental Research Funds for the Central Universities (2017CSY024, GK201603091, GK201703054)
Reference text:

彭亚丽,张鲁,张钰,刘侍刚,郭敏.基于深度反卷积神经网络的图像超分辨率算法.软件学报,2018,29(4):926-934

PENG Ya-Li,ZHANG Lu,ZHANG Yu,LIU Shi-Gang,GUO Min.Deep Deconvolution Neural Network for Image Super-Resolution.Journal of Software,2018,29(4):926-934