摘要本论文讨论了利用迭代法修复显微图像的算法,研究了图像复原算法在实际图像上的应用。首先对给定的显微图像进行了一系列分析,该图像为空间变化模糊,尺寸较大(1024×4096×3),边缘信噪比不高。在此基础上进行了一些探索,对不同种类的复原算法进行了模拟与比较。最终决定按空间不变方案来处理,采用非周期矩阵反卷积模型,使用受限制自适应复原算法实施修复,并辅助使用共轭梯度法、正则化和循环边界技术等加以完善。显微图像经过处理后,中心区域的图像清晰度显著提高,视觉效果得到了很大的改善,边缘区域也部分消除了模糊。近来获得了一些最新的显微图像,复原效果也非常明显。19438
关键词 反卷积 图像复原 共轭梯度法 空间自适应
毕业论文设计说明书(论文)外文摘要
Title Micrographs enhancement technology based on blind deconvolution
Abstract
This paper discusses the use of iterative restoration algorithms for micrographs and studied the application of image restoration algorithm to actual images. We analysed the given micrographs first, and found that their blur are space-variant and their size is large (1024×4096×3),at the same time, the signal to noise ratios of these micrographs are not high enough. We made some exploration on this basis,then we conducted simulation and comparison for different kinds of recovery algorithm. At last, we decided to deal with this problem by spatially invariant solutions. We constructed limited space adaptive restoration algorithm by the use of non-periodic matrix deconvolution model, conjugate gradient method, regularization and the cyclic boundary technology. After processing, the sharpness and visual effect of image center is significantly improved, and the edge area of micrograph is partly deblurred. We get some of the latest microscopes recently; the recovery effect is also satisfactory.
Keywords deconvolution image recovery conjugate gradient method space adaptive
目 次
1 绪论 2
1.1 研究的背景介绍 2
1.2 几种基本的图像复原方法 2
1.3 本文研究内容 3
2 图像复原数学基础及相关理论基础 4
2.1 卷积 4
2.2 二文离散卷积 5
2.3 图像退化模型 7
2.4 降晰函数模型 8
2.5 反卷积问题的病态 10
2.6 正则化 11
2.7 卷积方程的离散化与非周期矩阵反卷积模型 12
2.8 代数方程的迭代解法 14
3 实用图像复原算法 17
3.1 受限制自适应复原算法 17
3.2 受限制自适应复原算法和文纳滤波算法的比较 18
3.3 受限制自适应复原算法的改进 19
3.4 降晰矩阵分解法 20
4 图像分析与Matlab算法实现 23
4.1 显微图像分析 23
4.2 算法的Matlab实现 29
4.3 最新进展 31
结 论 34
致 谢 35
参考文献 36
1 绪论 基于盲反卷积的显微图像增强技术研究:http://www.youerw.com/tongxin/lunwen_10792.html