边界保持平滑滤波方法研究应用.docVIP

北京..大学毕业设计（论文） 图像降噪中边界保持平滑滤波方法研究 姓名： 学号： 指导老师： 摘要 当今社会是信息数字化时代,不管是学习,生活全部和信息数字紧密相关联,其中数字图像处理在其中占有着举足轻重地位。二十一世纪,数字图像处理技术高速发展,并广泛应用于识别领域,医学领域,体育领域等。 平滑滤波是图像处理学基础.最基础滤波方法是均值滤波和中值滤波,这两种滤波方法对噪声全部有抑制作用而且算法简单,不过造成图像变模糊尤其是边缘变模糊是无可避免。即使将滤波器加权后,效果有所改善,但理论是近似,所以效果仍不显著。 为了改善边缘模糊,我们发觉只要处理好灰度改变显著边缘,图像就会达成一个很好效果,要最大程度保持图片清楚,期望在进行平滑处理同时,检测出景物边界,然后对噪声进行处理。 本文用matlab编辑算法实现中值滤波,均值滤波,最小方差滤波,k近邻滤波对噪声处理,并进行性能分析和比较。 关键词:matlab;中值滤波;均值滤波;最小方差滤波;k近邻滤波 Abstact Modern society is digital , whether learning or life is closely associated with digital information, including digital image processing in which occupies a pivotal position .21 century, the rapid development of digital image processing technology widely used in the field of identification, medicalfication, the field of sports. The smoothing filter is the basis of image processing. The most basic filter is the mean filter and median filter, both filtering noise inhibited and the algorithm is simple, but lead to the edge of the image blurred without avoidable. Although the effect is improving ,the theory is approximate, so the effect is still not clear. In order to improve the edge blur, we found that handling the gray-scale variation significant edge, the image will reach a good effect, To the maximum extent to maintain the clarity of the picture, while performing smoothing processing, a scene boundary is detected, then, the noise is processing. In this paper, using matlab algorithm for editing median filtering, mean filtering, minimum variance filtering, k-nearest neighbor filtering noise processing and analysis and comparison of performance.Keywords: matlab; median filter; mean filter; minimum variance filtering; k nearest neighbor filter 目录 TOC \o 1-3 \h \z \u 摘要 I Abstact II 第1章:绪论 1 1.1 课题背景 1 1.2研究目标 1 1.3研究内容 1 1.4论文的组织安排 2 第2章 平滑滤波的相关知识 3 2.1噪声的相关知识 3 2.1.1 噪声的定义 3 2.1.2在matlab中添加噪声 4 2.2彩色图像的分解 5 2.3平滑的概念 6 2.3.1空间域方法 7 2.3.2频率域方法 7 2.3.3平滑算法 7 2.4峰