2重要基于Matlab的图像去噪算法的研究.doc

摘 要 在信息化的社会里，图像在信息传播中所起的作用越来越大。在许多情况下图像信息会受到各种各样噪声的影响，严重时会影响图像中的有用信息，所以，消除在图像采集和传输过程中而产生的噪声，保证图像受污染度最小，成了数字图像处理领域里的重要部分。 本文首先分析了图像增强技术相关知识，重点讨论了空间域滤波方法，然后本文主要研究分析邻域平均法、中值滤波法、维纳滤波法及模糊小波变换法的图像去噪算法，并进行相应的仿真。首先介绍图像处理应用时的常用函数及其用法；其次详细阐述了四种去噪算法原理及特点；最后运用Matlab软件对一张含噪图片（含高斯噪声或椒盐噪声）进行仿真去噪，通过分析仿真结果得出：均值滤波是典型的线性滤波，对高斯噪声抑制是比较好的；中值滤波是常用的非线性滤波方法，对椒盐噪声特别有效；维纳滤波对高斯噪声有明显的抑制作用；对小波系数进行阈值处理可以在小波变换域中去除低幅值的噪声和不期望的信号。 关键词：图像增强技术；空间域滤波；邻域平均法；中值滤波；维纳滤波；小波变换 Abstract In the information society, the image in the information transmission is used more and more widely. In many cases image’s information can be affected by various noises, seriously affect the useful information of a image，Therefore, ensuring the minimum of the noise and pollution in the process of image collection and transmission became an important part of the field. This paper first had an analysis of some related knowledge about image enhancement technology with emphasis on discussing the spatial domain methods. Then this paper mainly analysis and discuss the neighborhood average method, median filtering method, wiener filtering method and the fuzzy wavelet transform method of image denoising algorithm.，and the corresponding simulation.Firstly introduce the common image processing functions and its applications. Secondly elaborate the principles and characteristics of the four denoising algorithm. Finally using Matlab software to a noise images (including gaussian noise or salt pepper noise), and getting some conclusions from the simulation denoising analysis: average filtering is typical of linear filter, which is better used for gaussian noise. The median filter is a common nonlinear filtering method, especially effective to salt pepper noise. Through wiener filtering, the gaussian noise is inhibited obviously. Wavelet coefficients threshold processing in wavelet domain can remove noise and the the signal which is not expect. Key words：image enhancement technology ；spatial domain ；Average neighborhood；Median filter；W