遥感图像的MAP超分辨重建

刘涛; 钱锋; 张葆

doi:10.3788/YJYXS20183310.0884

您当前的位置：

首页 >

文章列表页 >

遥感图像的MAP超分辨重建

图像处理 | 更新时间：2020-07-06

- 遥感图像的MAP超分辨重建
- MAP super-resolution reconstruction of remote sensing image
- 液晶与显示 2018年33卷第10期页码：884-892
- 作者机构：
  
  1.中国科学院长春光学精密机械与物理研究所, 吉林长春 130033
  2.中国科学院大学, 北京 100049
- 作者简介：
  
  [ "刘涛(1993-), 男, 山东烟台人, 硕士研究生, 2016年于兰州大学物理学院获得理学学士学位, 主要从事机器视觉图像处理图像超分辨重建等方面研究。E-mail:m15526864901@163.com" ]
  [ "张葆(1966-), 男, 吉林省磐石人, 工学博士, 研究员, 博士生导师, 1989年、1994年于长春光机学院获得学士、硕士学位, 2004年于中国科学院长春光学机密机械与物理研究所获得博士学位, 主要研究方向为航空光电成像技术。E-mail:cleresky@vip.sina.com" ]
- 基金信息：
  
  国家自然科学基金(61705225);National Natural Science Foundation of China(61705225)
- DOI：10.3788/YJYXS20183310.0884
  中图分类号： TP394.1;TH691.9
- 收稿日期：2018-06-04，
  
  录用日期：2018-7-9，
  
  纸质出版日期：2018-10-05
- 稿件说明：
移动端阅览
刘涛, 钱锋, 张葆. 遥感图像的MAP超分辨重建[J]. 液晶与显示, 2018,33(10):884-892.

Tao LIU, Feng QIAN, Bao ZHANG. MAP super-resolution reconstruction of remote sensing image[J]. Chinese journal of liquid crystals and displays, 2018, 33(10): 884-892.
刘涛, 钱锋, 张葆. 遥感图像的MAP超分辨重建[J]. 液晶与显示, 2018,33(10):884-892. DOI： 10.3788/YJYXS20183310.0884.

Tao LIU, Feng QIAN, Bao ZHANG. MAP super-resolution reconstruction of remote sensing image[J]. Chinese journal of liquid crystals and displays, 2018, 33(10): 884-892. DOI： 10.3788/YJYXS20183310.0884.

摘要

遥感根本目的就是获得清晰的高空间分辨率的图像，从而可以进一步地分析处理。为了在遥感测量中获得更高空间分辨率、更高信噪比、更清晰的图像，本文对图像处理领域超分辨算法进行了研究。建立了一套拟合模拟现实的成像系统模型，在这种模型的基础之上，利用最大后验概率系统理论，讨论了现实情况中的运动模糊，噪声等情况，改进了MAP超分辨算法。实验结果表明：使用本文改进的基于MAP理论的Markov随机场约束的多帧超分辨重建算法，可以较好提高超分辨效果，与三次立方插值方法相比，PSNR至少提高约5.1 dB左右，与未改进的MAP方法相比，PSNR提高约0.2 dB左右。本文提出了动态的先验约束方法，给约束函数添加与迭代次数相关的约束项，该改进创新可以加快收敛并且更加逼近真实图像，实验表明该方法收敛速度更快，约束效果良好，更适合实际应用。

Abstract

The fundamental purpose of remote sensing is to obtain clear images with high spatial resolution

which can be further analyzed and processed. In order to obtain higher spatial resolution

higher signal-to-noise ratio and clearer images in remote sensing measurement

this paper studies the superresolution algorithm in image processing field. In the paper we set up a model fitting the real imaging system

use the theory of maximum a posteriori probability system on this model

discuss the motion blur

noise and other degrading factors

and finally improve the MAP superresolution algorithm. The experimental results show that the improved multi-frame superresolution reconstruction algorithm in the paper based on the Markov constraint field which is originated from the MAP theory

can improve the superresolution performance significantly. Using our method

PSNR is at least about 5.1 dB higher than the cubic interpolation method

and is about 0.2 dB higher than the unimproved MAP method. In our paper a dynamic prior constraint method is proposed

by adding constraints associated with the number of iterations to constraint function

the improvement innovation can accelerate convergence and more close to real images. And the experiments show that this method can fasten the convergence

and its constraint effect is good. Above all

it is more suitable for practical application.

关键词

Keywords

references

何林阳.航空图像超分辨率重建关键技术研究[D].长春: 中国科学院大学(长春光学精密机械与物理研究所), 2016. http://cdmd.cnki.com.cn/Article/CDMD-80139-1016082576.htm

HE L Y. Research on key techniques of super-resolution reconstruction of aerial images[D].Changchun: University of the Chinese Academy of Sciences (CIOMPS), 2016. (in Chinese)

WANG T W, SUN W J, QI H R, et al. Aerial Image super resolution via wavelet multiscale convolutional neural networks[J]. IEEE Geoscience and Remote Sensing Letters, 2018, doi:10.1109/LGRS.2018.2810893.

REN C, HE X H, PU Y F. Nonlocal similarity modeling and deep CNN gradient prior for super resolution[J]. IEEE Signal Processing Letters, 2018, 25(7):916-920.

LIU D, WANG Z W, FAN Y C, et al. Learning temporal dynamics for video super-resolution:a deep learning approach[J]. IEEE Transactions on Image Processing, 2018, 27(7):3432-3445.

REZAYI H, SEYEDIN S A. Huber Markov random field for joint super resolution[C]//2017 10 th Iranian Conference on Machine Vision and Image Processing ( MVIP ). Isfahan, Iran: Isfahan University of Technology, 2017.

BAO W B, ZHANG X Y, YAN S P, et al . Iterative convolutional neural network for noisy image super-resolution[C]//2017 IEEE International Conference on Image Processing ( ICIP ). Beijing, China: IEEE, 2017.

ZHA Z Y, LIU H, LI J K. Based on the technique of regularization MAP super-resolution image reconstruction algorithm[C]//2014 Seventh International Symposium on Computational Intelligence and Design . Washington, DC, USA: IEEE Computer Society, 2014.

ZHANG H Y, ZHANG L P, SHEN H F, et al . A MAP approach for joint image registration, blur identification and super resolution[C]//2009 Fifth International Conference on Image and Graphics .Xi'an: IEEE, 2009.

HÄFNER M, LIEDLGRUBER M, UHL A. POCS-based super-resolution for HD endoscopy video frames[C]// Proceedings of the 26 th IEEE International Symposium on Computer-Based Medical Systems . Porto, Portugal: IEEE, 2013.

朱齐丹, 孙磊, 蔡成涛.多正则化形式的超分辨率图像重建[J].光电工程, 2015, 42(1):45-50.

ZHU Q D, SUN L, CAI C T. Image super-resolution with multiple regularized terms [J]. Opto-Electronic Engineering, 2015, 42(1):45-50. (in Chinese)

刘颖, 权婉, 伍世虔.基于边缘增强的正则化超分辨率图像重建[J].西安邮电大学学报, 2016, 21(6):14-19.

LIU Y, QUAN W, WU S Q. Regularized super-resolution image reconstruction based on edge enhancement[J]. Journal of Xi'an University of Posts and Telecommunications, 2016, 21(6):14-19. (in Chinese)

韩玉兰, 赵永平, 王启松, 等.稀疏表示下的噪声图像超分辨率重构[J].光学精密工程, 2017, 25(6):1619-1626.

HAN Y L, ZHAO Y P, WANG Q S, et al. Sparse representation of noise image super-resolution reconstruction[J]. Optics and Precision Engineering, 2017, 25(6):1619-1626. (in Chinese)

孙刚.基于MAP的红外图像超分辨率重建算法研究[D].西安: 西安电子科技大学, 2008. http://cdmd.cnki.com.cn/article/cdmd-10701-2008056050.htm

SUN G. Research on the algirithm of super-resolution reconstruction of infrared images based on MAP[D]. Xi'an: Xidian University, 2008. (in Chinese)

穆绍硕.基于多帧影像的航空超分辨成像技术研究[D].长春: 中国科学院大学(中国科学院长春光学精密机械与物理研究所), 2017. http://cdmd.cnki.com.cn/Article/CDMD-80139-1017063984.htm

MU S S. Research on super-resolution imaging techniques of multi-frame aerial images[D]. Changchun: University of the Chinese Academy of Sciences (CIOMPS), 2017. (in Chinese)

康凯.图像超分辨率重建研究[D].合肥: 中国科技大学, 2016.

KANG K. A study on image super-resolution reconstruction[D]. Hefei: University of Science and Technology of China, 2016. (in Chinese)

孙辉, 马天玮.基于相位相关的目标图像亚像元运动参数估计[J].液晶与显示, 2011, 26(6):858-862.

SUN H, MA T W. Sub-pixel motion estimation based on phase-only correlation [J]. Chinese Journal of Liquid Crystals and Displays, 2011, 26(6):858-862. (in Chinese)

WEN R, FU K, SUN H, et al. Image superresolution using densely connected residual networks[J]. IEEE Signal Processing Letters. 2018;25(10):1565-1569.

RASTI P, DEMIREL H, ANBARJAFARI G. Iterative back projection based image resolution enhancement[C]//2013 8 th Iranian Conference on Machine Vision and Image Processing ( MVIP ). Zanjan, Iran: IEEE, 2013.

IYANDA C A, YAAKOB S N B, NAWIR M. Uniqueness of iterative back projection in super resolution techniques[C]//2016 3 rd International Conference on Electronic Design ( ICED ). Phuket, Thailand: IEEE, 2016.

谭周燚, 黄自力, 王雪梅, 等.基于结构自适应归一化卷积的超分辨率红外图像重建[J].电子技术, 2017, 46(2):13-20.

TAN Z Y, HUANG Z L, WANG X M, et al.Infrared image super-resolution based on the structure-adaptive normalized convolution[J]. Electronic Technology, 2017, 46(2):13-20. (in Chinese)

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据