[1]史永祥,蒋斌,黄雍晫,等.基于深度学习的红外图像超分辨率重建[J].应用科技,2020,47(4):8-13.[doi:10.11991/yykj.201912020]
　SHI Yongxiang,JIANG Bin,HUANG Yongzhuo,et al.Infrared image super-resolution reconstruction based on deep learning[J].Applied science and technology,2020,47(4):8-13.[doi:10.11991/yykj.201912020]
点击复制

基于深度学习的红外图像超分辨率重建(/HTML)
分享到：

《应用科技》[ISSN:1009-671X/CN:23-1191/U]

卷:
第47卷
期数:
2020年4期
页码:
8-13
栏目:
智能科学与技术
出版日期:
2020-07-05

文章信息/Info

Title:
Infrared image super-resolution reconstruction based on deep learning
作者:
史永祥1;  蒋斌1;  黄雍晫1;  杨桂生1;  李庆武2;  张志良2
1. 国家电网溧阳市供电公司，江苏 溧阳 213300;
2. 河海大学 物联网工程学院，江苏 常州 213000
Author(s):
SHI Yongxiang1;  JIANG Bin1;  HUANG Yongzhuo1;  YANG Guisheng1;  LI Qingwu2;  ZHANG Zhiliang2
1. State Grid Liyang Power Supply Company, Liyang 213300, China;
2. School of Internet of Things Engineering, Hohai University, Changzhou 213000, China
关键词:
神经网络; 深度学习; 残差网络; 红外图像; 超分辨率重建; 池化层; 感受野; 增强预测
Keywords:
neural network; deep learning; residual network; infrared image; super-resolution reconstruction; pool layer; receptive field; enhanced forecasting
分类号:
TP31
DOI:
10.11991/yykj.201912020
文献标志码:
A
摘要:
为提升红外图像分辨率，本文构建了用于红外图像超分辨率重建的IEDSR（enhanced deep residual networks for infrared image super-resolution）网络。该网络在EDSR网络模型的基础上加入了池化层，避免了EDSR（enhanced deep residual networks for single image super-sesolution）网络移除批正则化层（batch normalization, BN）可能会带来训练困难的问题。同时考虑到红外图像对比度低、纹理不明显的特性，在残差块内加入新的卷积层和激活层，通过增加网络深度扩大局部残差模块的感受野，有利于恢复图像的局部细节信息。最后利用增强预测算法对重建图像进行优化，提升重建精度。实验结果表明：本文算法重建的红外图像在主观视觉效果与客观指标上较传统红外图像重建方法均有所改善，具有较高的实用价值。
Abstract:
In order to improve the resolution of infrared image, this paper constructs an IEDSR (enhanced deep residual networks for infrared image super-resolution) network for infrared image super-resolution reconstruction. Based on the EDSR (enhanced deep residual networks for single image super-sesolution) network model, a pooling layer is added to the network, which avoids the problem that removing BN (batch normalization) layer from EDSR network may bring training difficulty. At the same time, considering the low contrast of infrared image and the characteristics of not obvious texture, a new convolution layer and activation layer are added to the residual block, which is helpful to recover the local details of the image by increasing the depth of network and expanding the receptive field of the local residual module. Finally, we use the enhanced prediction algorithm to optimize the reconstructed image and improve the reconstruction accuracy. The experimental results show that the infrared image reconstructed by this algorithm has better subjective visual effect and objective index than traditional infrared image reconstruction method, and has higher practical value.

参考文献/References:

[1] 王合龙, 边栓成. 一种增强细节的红外图像处理算法[J]. 太赫兹科学与电子信息学报, 2018, 16(1): 139–142
[2] 王岳, 李双喜, 王磊. 红外图像超分辨率重建技术研究[J]. 激光与红外, 2018, 48(4): 524–530
[3] 苏衡, 周杰, 张志浩. 超分辨率图像重建方法综述[J]. 自动化学报, 2013, 39(8): 1202–1213
[4] 李浪宇, 苏卓, 石晓红, 等. 图像超分辨率重建中的细节互补卷积模型[J]. 中国图象图形学报, 2018, 23(4): 572–582
[5] TIAN Jing, MA Kaikuang. A survey on super-resolution imaging[J]. Signal, image and video processing, 2011, 5(3): 329–342.
[6] 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
[7] 廖小华, 陈念年, 蒋勇, 等. 改进的卷积神经网络红外图像超分辨率算法[J]. 红外技术, 2020, 42(1): 75–80
[8] DONG C, LOY C, HE K, et al. Learning a deep convolutional network for image super-resolution[C]//ECCV. Zurich, Switzerland，2014: 35-40.
[9] 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 (CVPR). Honolulu, Hawaii, 2017：105-114.
[10] HE Kaiming, ZHANG Xiangyu，REN Shaoqing, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, USA, 2016: 770-778.
[11] GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]//Advances in neural information processing systems. Montreal, Canada, 2014: 2672-2680.
[12] LIM B, SON S, KIM H, et al. Enhanced deep residual networks for single image super-resolution[C]//Computer Vision and Pattern Recognition Workshops. Honolulu, USA, IEEE, 2017: 1132-1140.
[13] SHAHAM T R, DEKEL T, MICHAELI T. SinGAN: Learning a generative model from a single natural image[C]//2019 IEEE/CVF International Conference on Computer Vision (ICCV). Seoul, Korea , 2019: 4569-4579.
[14] 凡遵林, 管乃洋, 王之元, 等. 红外图像质量的提升技术综述[J]. 红外技术, 2019, 41(10): 941–946
[15] 张川. 面向图像分类的深度残差网络优化结构研究[D]. 北京：中国科学院大学，2016.
[16] IOFFE S, SZEGEDY C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]//International Conference on International Conference on Machine Learning. Lille, French, 2015: 448-456.
[17] 麻旋, 戴曙光. 基于残差网络的图像超分辨率算法改进研究[J]. 软件导刊, 2018, 17(4): 95–97
[18] TIMOFTE R, ROTHE R, VAN GOOL L. Seven ways to improve example-based single image super resolution[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, USA, 2016: 1865-1873.
[19] 王贺, 李野, 付明艳. 一种基于图像分层的红外图像目标细节增强算法[J]. 计算机科学与应用, 2020, 10(5): 6
[20] BOOR C D. Bicubic spline interpolation[J]. Journal of mathematics and physics, 1962, 41(1/2/3/4):212-218.

相似文献/References:

[1]王琦,苏畅,姜弢.“巨”型负磁导率材料的优化设计[J].应用科技,2011,38(04):19.[doi:doi:10.3969/j.issn.1009-671X.2011.04.05]
　WANG Qi,SU Chang,JIANG Tao.Optimal design of negative permeability material with “巨”[J].Applied science and technology,2011,38(4):19.[doi:doi:10.3969/j.issn.1009-671X.2011.04.05]
[2]刘少刚,吕建伟,王士成.基于数据融合与神经网络的火灾探测研究[J].应用科技,2011,38(05):9.[doi:10.3969/j.issn.1009-671X.2011.05.03]
　LIU Shaogang,LV Jianwei,WANG Shicheng.A primitive study of fire detection method based on data fusion and BP neural network[J].Applied science and technology,2011,38(4):9.[doi:10.3969/j.issn.1009-671X.2011.05.03]
[3]张文辉,胡小平,朱银发.自由漂浮空间机械臂基于神经网络的H∞鲁棒控制[J].应用科技,2012,39(06):5.[doi:10.3969/j.issn.1009-671X.2012.201209005]
　ZHANG Wenhui,HU Xiaoping,ZHU Yinfa.Neural network based H∞ robust control of free-floating space robot manipulators[J].Applied science and technology,2012,39(4):5.[doi:10.3969/j.issn.1009-671X.2012.201209005]
[4]袁建东,夏国清.神经网络监督控制在船用一体化压水堆功率控制中的应用[J].应用科技,2005,32(01):24.
　YUAN Jiandong,XIA Guoqing.Application of neural network supervisory control to power regulating of marine integral pressurized water reactor[J].Applied science and technology,2005,32(4):24.
[5]廖艳苹,谢 红,杨莘元.辅助式小波神经网络的调制识别技术[J].应用科技,2006,33(04):24.
　LIAO Yan-ping,XIE Hong,YANG Shen-yuan.Recognition of modulation using assistant wavelet neural network[J].Applied science and technology,2006,33(4):24.
[6]侯艳雪,黄曼磊,陶丽楠,等.神经网络在永磁同步电机变频调速系统中的应用[J].应用科技,2015,42(02):1.[doi:10.3969/j.issn.1009-671X.201405020]
　HOU Yanxue,HUANG Manlei,TAO Linan,et al.Application of artificial neural network (ANN) in variable frequency speed control system of PMSM[J].Applied science and technology,2015,42(4):1.[doi:10.3969/j.issn.1009-671X.201405020]
[7]陈立伟,黄璐,齐传斌.基于遗传算法优化的相关向量机的燃机涡轮叶片故障诊断[J].应用科技,2016,43(02):70.[doi:10.11991/yykj.201506026]
　CHEN Liwei,HUANG Lu,QI Chuanbin.Fault diagnosis for gas turbine blades based on the relevance vector machine optimized by genetic algorithm[J].Applied science and technology,2016,43(4):70.[doi:10.11991/yykj.201506026]
[8]丁虎,姚磊,刘少刚,等.基于神经网络和图像分割的林火图像识别研究[J].应用科技,2016,43(03):82.[doi:10.11991/yykj.201510011]
　DING Hu,YAO Lei,LIU Shaogang,et al.The forest fire image recognition based on neural network and image segmentation[J].Applied science and technology,2016,43(4):82.[doi:10.11991/yykj.201510011]
[9]朱少民,夏虹,杨波,等.蒸发器水位控制系统在线故障诊断方法[J].应用科技,2016,43(05):75.[doi:10.11991/yykj.201512017]
　ZHU Shaomin,XIA Hong,YANG Bo,et al.On-line fault diagnosis method of steam generator water level control system[J].Applied science and technology,2016,43(4):75.[doi:10.11991/yykj.201512017]
[10]张秋云,张营,李臣.遗传算法优化BP神经网络在中医按摩机器人中的应用[J].应用科技,2017,44(02):73.[doi:10.11991/yykj.201603018]
　ZHANG Qiuyun,ZHANG Ying,LI Chen.Application of BP neural network based on genetic algorithm optimization in Chinese medicine massage robot[J].Applied science and technology,2017,44(4):73.[doi:10.11991/yykj.201603018]
[11]卢用煌,黄山.深度学习在身份证号码识别中的应用[J].应用科技,2019,46(01):123.[doi:10.11991/yykj.201804008]
　LU Yonghuang,HUANG Shan.Application of deep learning in identification of ID card number[J].Applied science and technology,2019,46(4):123.[doi:10.11991/yykj.201804008]
[12]李茜萌,陈春雨,何恒翔.弹性部署的分布式AI计算架构系统研究[J].应用科技,2020,47(5):64.[doi:10.11991/yykj.202006001]
　LI Ximeng,CHEN Chunyu,HE Hengxiang.Research on distributed AI computing architecture system with flexible deployment[J].Applied science and technology,2020,47(4):64.[doi:10.11991/yykj.202006001]

备注/Memo

备注/Memo:
收稿日期:2019-12-20。
基金项目:国家自然科学基金项目(41706103)；国家重点研发计划项目(2018YFC0406900)
作者简介:史永祥，男，高级工程师;蒋斌，男，高级工程师
通讯作者:张志良，E-mail：xjolyon@gmail.co
更新日期/Last Update: 2020-11-27