中国科学院光电技术研究所机构知识库
Advanced  
IOE OpenIR  > 光电技术研究所博硕士论文  > 学位论文
题名:
光电成像系统图像复原算法的研究
作者: 祝豪
学位类别: 博士
答辩日期: 2008-06-06
授予单位: 中国科学院光电技术研究所
授予地点: 光电技术研究所
导师: 吴钦章
关键词: 盲反卷积 ; 极大似然估计 ; 湍流退化 ; 图像复原
其他题名: Research on Image Restoration Algorithms in Photoelectric Imaging System
学位专业: 信号与信息处理
中文摘要: 由于实际物理条件的限制,通常观测到的图像都存在着各种不同程度的退化模糊。为了有效识别目标,必须先对模糊图像进行预处理以恢复出清晰的图像。图像复原的目的就是从观测到的退化图像中恢复出清晰的目标图像。本文重点研究了如何针对光电成像系统中的大气湍流退化图像进行有效复原,主要涉及湍流退化图像盲反卷积算法的研究。 根据点扩散函数的概率密度特性,假设某一类退化图像的点扩散函数为G类点扩散函数,采用APEX方法和非线性最小二乘拟合算法,可以得到点扩散函数解析表达式,之后采用SECB算法根据估计出的点扩散函数进行图像复原。为了克服SECB算法复原出的图像带有明显的噪声和振铃效应的缺陷,提出了一种具有保真项和高阶凸正则化项的各向异性扩散模型,有效的抑制了噪声和振铃效应。 假设观测图像服从Poisson分布,基于极大似然理论,根据Picard迭代方法,提出了一种单帧Poisson指数迭代复原算法,并将其推广,得到了一种收敛速度更快的加速Poisson指数迭代算法,该算法具有良好的图像复原性能。 对于传统的多帧图像期望最大化复原算法,随着迭代次数的增加,恢复出的目标图像容易带来噪声甚至出现无效解,且恢复出的点扩散函数的带宽超出了光电成像系统的带宽。针对这些缺点,本文提出了一种具有带宽和总变分约束的盲复原算法。将总变分约束加入到似然函数中用以抑制噪声,并通过带宽有限函数来控制点扩散函数的带宽在成像系统的有效带宽范围之内。该算法能有效抑制噪声,避免了恢复的目标图像出现无效解,实验结果证明了该算法的优越性。 在极大似然估计理论的基础上,将观测图像所满足的似然函数的最大化问题转化成目标函数的极小化进行求解。根据实际先验信息,即目标图像和点扩散函数的非负性以及点扩散函数的带宽有限性,建立了一种具有强制性非负约束和点扩散函数带宽惩罚函数的优化目标函数。根据最优化理论,采用共轭梯度优化算法,通过区间限定、黄金分割搜索以及迭代二次逼近的混合最优步长搜索方法对优化目标函数进行寻优求解。实验结果证明了该梯度优化算法的有效性以及在噪声情况下的鲁棒性。 在单帧图像Poisson指数迭代算法的基础上,提出了多帧图像的Poisson指数迭代算法,与单帧Poisson指数迭代算法相比,多帧情况下的复原算法能够获得更加清晰的复原图像。由于在分辨率最高时,点扩散函数将是极狭窄的,而且没有旁瓣的,将旁瓣抑制引入到多帧图像Poisson指数迭代算法中。通过构造旁瓣抑制函数,在迭代过程对点扩散函数进行旁瓣抑制,保证恢复出的目标图像更加接近真实解。实验表明,旁瓣抑制多帧Poisson指数迭代算法恢复出的目标图像具有更明显的细节纹理特征和具有更高的分辨率。
英文摘要: Due to the limitation of practical physical conditions, the observed images are usually degraded with different degrees. In order to identify the object effectively, it is necessary to restore the clear images through the pretreatment of the blurred images. The purpose of the image restoration is to restore the vivid images from the observed degraded images. The thesis focuses on how to effectively restore the degraded images of the photo-electricity imaging system owing to the influence of the atmospheric turbulence, and is mostly involved in the research of blind deconvolution algorithm of turbulence degraded images. Thanks to the probability density characteristic of the point spread function(PSF), on the assumption that the PSFs of some degraded images are class G PSF, the APEX method and the nonlinear least-squares algorithm are used to obtain the analysis expression of the PSF. Then, according to the estimated PSF, the SECB algorithm is performed to restore the images. To overcome the shortcoming that the restored images of the SECB algorithm have obvious noise and ringing effect, an anisotropic diffusion model with the fidelity term and the convex regularization term is introduced, which can restrain the noise and ringing effect effectively. Suppose that the observed images obey Poisson distribution, on the basis of the maximum likelihood theory, according to Picard iteration method, a single-frame Poisson exponent iterative (PEI) restoration algorithm is put forward. Then the PEI algorithm is generalized and further an accelerated PEI (APEI) algorithm is obtained, which is with a faster convergence speed and a favorable restoration capability. As to the traditional multi-frame Expectation-Maximization (EM) image restoration algorithm, with the increasing number of iterations, the restored images have noise and even appear trivialism solution, and the bandwidth of the estimated PSFs exceeds the bandwidth of the photo-electricity imaging system. Aimed at these disadvantages, a blind image deconvolution algorithm with bandwidth and total variation (TV) constraints is proposed. Joining the TV constraint with likelihood function, the algorithm can control noise, and depending on the band limited function, the bandwidth of the PSF is restricted within the range of that of the imaging system. This algorithm can remove noise effectively and avoid the trivialism solution, and the experimental results demonstrate the superiority of the algorithm. Based on the maximum likelihood estimate theory, the maximum problem of the likelihood function with which the observed images are satisfied turns into the minimum problem of the object function. According to the practical prior information, namely the object image and the PSFs are nonnegative and the limitation of the bandwidth of the PSFs, an optimized object function with compellent nonnegative constraint and the bandwidth punished function is constructed. Due to the optimization theory, by using of the conjugate gradient optimization algorithm, and through the optimization step length search method comprising the interval qualification, the golden section search and the iterative quadratic approximation method, the optimized object function is performed. The experimental results prove the efficacious of the gradient optimization algorithm and the robust under the noise condition. A multi-frame Poisson exponent iterative (MPEI) image restoration algorithm is brought forward based on the PEI algorithm. Compared with the PEI algorithm, the much clearer images can be achieved by the MPEI algorithm. Since the PSF would be extremely narrow and without side-lobes for best resolution, the side-lobe suppression is brought in the MPEI algorithm. Through constructing the side-lobe suppression function (SSF), the PSF is constraint by the SSF in the iteration and ensure the restored image is much close to the truth solution. The experimental results illustrate that the restored images using the side-lobe suppression MPEI algorithm have more obviously texture character and the higher resolution.
语种: 中文
内容类型: 学位论文
URI标识: http://ir.ioe.ac.cn/handle/181551/297
Appears in Collections:光电技术研究所博硕士论文_学位论文

Files in This Item:
File Name/ File Size Content Type Version Access License
10001_200518015127036祝豪_paper.pdf(3398KB)----限制开放View 联系获取全文

Recommended Citation:
祝豪. 光电成像系统图像复原算法的研究[D]. 光电技术研究所. 中国科学院光电技术研究所. 2008.
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[祝豪]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[祝豪]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
文件名: 10001_200518015127036祝豪_paper.pdf
格式: Adobe PDF
此文件暂不支持浏览
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2016  中国科学院光电技术研究所 - Feedback
Powered by CSpace