基于PSF重构和改进的最大后验估计的自适应光学图像复原算法

武兴睿

doi:10.3788/YJYXS20193409.0921

您当前的位置：

首页 >

文章列表页 >

基于PSF重构和改进的最大后验估计的自适应光学图像复原算法

图像处理 | 更新时间：2020-07-05

- 基于PSF重构和改进的最大后验估计的自适应光学图像复原算法
- Adaptive optical image restoration method based on PSF reconstruction and improved Maximum A Posteriori estimation
- 液晶与显示 2019年34卷第9期页码：921-927
- 作者机构：
  
  吉林科技职业技术学院, 吉林长春 130123
- 作者简介：
  
  [ "武兴睿(1982—), 女, 吉林长春人, 讲师, 2005年于长春大学获得学士学位, 主要从事计算机大数据技术、图形图像方面的研究。E-mail:jltg2019@163.com" ]
- 基金信息：
  
  吉林省高教学会重点项目(JGJX2018C106);Higher Education Society Key Project of Jilin Provice(JGJX2018C106)
- DOI：10.3788/YJYXS20193409.0921
  中图分类号： TP391.4
- 收稿日期：2019-04-12，
  
  录用日期：2019-5-20，
  
  纸质出版日期：2019-09-05
- 稿件说明：
移动端阅览
武兴睿. 基于PSF重构和改进的最大后验估计的自适应光学图像复原算法[J]. 液晶与显示, 2019,34(9):921-927.

Xing-rui WU. Adaptive optical image restoration method based on PSF reconstruction and improved Maximum A Posteriori estimation[J]. Chinese journal of liquid crystals and displays, 2019, 34(9): 921-927.
武兴睿. 基于PSF重构和改进的最大后验估计的自适应光学图像复原算法[J]. 液晶与显示, 2019,34(9):921-927. DOI： 10.3788/YJYXS20193409.0921.

Xing-rui WU. Adaptive optical image restoration method based on PSF reconstruction and improved Maximum A Posteriori estimation[J]. Chinese journal of liquid crystals and displays, 2019, 34(9): 921-927. DOI： 10.3788/YJYXS20193409.0921.

摘要

自适应光学（AO）系统为大气湍流提供实时补偿，但AO图像的对比度通常很差，这会导致图像质量下降。本文提出了一种基于点扩散函数（PSF）重构和改进的最大后验概率（MAP）估计的自适应光学图像恢复方法。首先，结合观测条件和AO系统的特点，建立基于波前相位信息的PSF重构模型；其次，基于本文提出的算法，建立了AO图像的迭代求解公式，实现多帧AO图像联合去卷积过程。为了验证本文算法的有效性，对仿真的退化AO图像进行了一系列复原实验。实验结果表明，以“Man”图像为例，与Wiener-IBD、RL-IBD及FS-MLJD算法相比，该算法的峰值信噪比（PSNR）测度分别提高了6.83%，4.47%和2.28%，拉普拉斯梯度模（LS）值分别提高了22.2%，17.9%和12.1%。本文的研究结果对实际的AO图像恢复具有一定的应用价值。

Abstract

The adaptive optics (AO) system can provide real-time compensation for atmospheric turbulence. However

an AO image is usually of poor contrast and brings about a loss in quality. In this paper

a adaptive optical image restoration method was presented

via

the Point Spread Function (PSF) reconstruction and improved Maximum A Posteriori (MAP) estimation. Firstly

combining the observed conditions and AO system characteristics

a predicted PSF model for the wavefront phase effect was developed. Secondly

the iterative solution formulas of AO image was build up based on the proposed algorithm

addressing the implementation process of multi-frame AO images joint deconvolution method. A series of experiments on simulated degraded AO images were conducted to evaluate our proposed algorithm. Compared with the Wiener iterative blind deconvolution (Wiener-IBD) algorithm

Richardson-Lucy IBD algorithm and FS-MLJD algorithm

our algorithm has better restoration effects including higher Peak Signal-to-Noise Ratio (PSNR) and Laplacian Sum (LS) values than the others. For "Man" image

the PSNR has improved 6.83%

4.47% and 2.28%

and the LS value has improved 22.2%

7.9% and 12.1%

respectively. The research results have a certain application values for actual AO image restoration.

关键词

Keywords

references

FAROOQ M, ASLAM A, HUSSAIN B, et al . A comparison of image processing techniques for optical interference fringe analysis[J]. Photonic Sensors , 2015, 5(4): 304-311.

AYERS G R, DAINTY J C. Iterative blind deconvolution method and its applications[J]. Optics Letters , 1998, 13(7): 547-549.

TIAN Y, RAO C H, WEI K. Adaptive optics image restoration based on frame selection and multi-Frame blind deconvolution[J]. Chinese Astronomy and Astrophysics , 2009, 33(2): 223-230.

张丽娟, 杨进华, 苏伟.基于改进的期望值最大化自适应光学图像多帧迭代去卷积算法[J].兵工学报, 2014, 35(11):1765-1773.

ZHANG L J, YANG J H, SU W. Multi-frame iteration blind deconvolution algorithm based on improved expectation maximization for adaptive optics image restoration[J]. Acta Armamentarii , 2014, 35(11): 1765-1773. (in Chinese).

YAP K H, GUAN L. Adaptive image restoration based on hierarchical neural networks[J]. Optical Engineering , 2000, 39(7): 1877-1890.

饶长辉, 沈锋, 姜文汉.自适应光学系统波前校正残余误差的功率谱分析方法[J].光学学报, 2000, 20(1):68-73.

RAO C H, SHEN F, JIANG W H. Analysis of closed-Loop wavefront residual error of adaptive optical system using the method of power spectrum[J]. Acta Optica Sinica , 2000, 20(1): 68-73 (in Chinese).

陈波.自适应光学图像复原理论与算法研究[D].郑州: 解放军信息工程大学, 2008: 44-46.

CHEN B. The theory and algorithms of adaptive optics image restoration [D]. Zhengzhou: Information Engineering University, 2008: 44-46. (in Chinese)

HUSSAIN B, MUHAMMAD T, REHAN M, et al . Fast processing of optical fringe movement in displacement sensors without using an ADC[J] . Photonic Sensors , 2013, 3(3): 241-245.

LI D M, SU Z B, SU W, et al . Research on cross-correlative blur length estimation algorithm in motion blur image[J]. Journal of Advanced Computational Intelligence and Intelligent Informatics , 2016, 20(1): 155-162.

TANG X J. Regularization method for image restoration [D]. Wuhan: Huazhong University, 2006: 101-113.

ALLAN W.The USC-SIPI image database[EB/OL].[2019-03-10] . http://sipi.usc.edu/database/ http://sipi.usc.edu/database/ .

TSUMURAYA F, MIURA N, BABA N. Iterative blind deconvolution method using Lucy's algorithm [J]. Astronomy and Astrophysics , 1994, 282(2): 699-708.

张士杰, 李俊山, 杨亚威, 等.湍流退化红外图像降晰函数辨识[J].光学精密工程, 2013, 21(2):514-521.

ZHANG S J, LI J S, YANG Y W, et al . Blur identification of turbulence-degraded IR images[J]. Optics and Precision Engineering , 2013, 21(2): 514-521 (in Chinese).

张丽娟, 李东明, 杨进华, 等.基于帧选择和极大似然估计的自适应光学图像多帧联合去卷积算法[J].吉林大学学报:理学版, 2017, 55(5):1199-1206.

ZHANG L J, LI D M, YANG J H, et al . Multi-frame joint deconvolution algorithm for adaptive optics images based on frame-selection and maximum likelihood estimation[J]. Journal of Jilin University : Science Edition , 2017, 55(5): 1199-1206. (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

基于深度残差生成对抗网络的运动图像去模糊

基于多尺度融合卷积神经网络的图像去雾算法

基于YUV颜色模型与导向滤波的图像去雾算法

结合双阈值定位与透射率约束的航拍图像去雾

基于多尺度融合卷积神经网络的图像去雾算法