摘要：数字图像在采集、传输和存储的过程中可能会受各种因素的影响而导致图像质量下降,这种现象称为图像“退化”。我们可以利用图像复原技术由退化的图像得到清晰的图像。在对目前有关数字图像复原技术的文献进行理解和综合的基础上，本论文主要介绍了基于Matlab图像复原技术的研究背景、现状以及发展方向，阐述了采用Matlab图像处理工具箱中来实现图像复原的方法。
本论文实现了四个图像复原算法，即文纳滤波算法、约束最小二乘方（正则）复原算法、迭代非线性复原算法（Lucy-Richardson算法）以及盲去卷积算法。对它们的原理及相应的Matlab函数进行了介绍，然后对不同场景的图像进行仿真，并分析了实验的结果。根据仿真结果我们可以很清楚的看出，在已知退化函数的情况下，这四种算法中采用文纳滤波算法得到的复原图像比较清晰，盲去卷积算法得到的复原图像质量较差；而在退化函数未知的情况下，用盲去算法得到的复原图像质量最好。8829
关键词： Matlab；文纳滤波；约束最小二乘方复原；迭代非线性复原；盲去卷积
Study on the Recovery Technology of Railway Monitoring Image
Abstract: Digital images in the collection, transmission and storage process may be affected by various factors which led to decline in image quality, a phenomenon known as Image "degraded." We can use an image from the degraded image restoration technique to get a clear image. In the current technology for digital image recovery and synthesis of the literature on the basis of understanding, this paper introduces the Matlab-based image restoration technology research background, current status and direction of development, elaborated in using Matlab Processing Toolbox to achieve image restoration approach.
This paper implements four image restoration algorithm, namely Wiener Filter algorithm; Constrained least squares(Regular) restoration algorithm; Iterative non-linear restoration algorithm(Lucy-Richardson algorithm) and blind deconvolution algorithms. For their principles and the corresponding Matlab function were introduced, and then the images of different simulation scenarios and analyze the experimental results. According to the simulation results, we can clearly be seen in the case of the known degenerate function, of these four algorithms, wiener filtering algorithm, which used to recover the image is relatively clear, blind deconvolution algorithm to recover the image quality is poor; in the degenerate case of unknown function, a restoration blind deconvolution algorithms is the best image quality.
Key words: Matlab; Wiener Filter; Constrained least squares recovery; Iterative non-linear recovery; Blind deconvolution
目 录
1绪论 1
1.1 研究背景与研究意义 1
1.2 国内外研究现状与水平 2
1.3 发展趋势 2
1.4 本文的主要工作及内容安排 4
2图像的退化 5
2.1 图像退化模型 5
2.2 图像噪声 6
2.2.1 图像噪声的概念 6
2.2.2 图像噪声的分类 6
2.3用函数复原图像 7
3几种经典的复原方法介绍 8
3.1 图像复原的本质 8
3.2 文纳滤波复原法 8
3.3 约束最小二乘方（正则）复原算法 9
3.4 迭代非线性复原法（Lucy-Richardson法） 10
3.5 盲去卷积法 12
4实验结果与分析 13
4.1 图像质量评价标准 15
4.2 利用文纳滤波进行复原 16
4.3 利用约束最小二乘方复原法进行复原 17 Matlab铁路监控图像的复原技术研究:http://www.751com.cn/tongxin/lunwen_7310.html