《数字图像的去噪算法研究与实现毕业论文说明书》.docVIP

LANZHOU UNIVERSITY OF TECHNOLOGY 毕业论文 题 目 数字图像的去噪算法研究与实现 学生姓名 ** 学 号 ** 专业班级 通信工*班 指导教师 ** 学 院 计算机与通信学院 答辩日期 *年6月16日 数字图像的去噪算法研究与实现 Denoising Algorithm and Implementation for Digital Image * * * 摘要 数字图像的去噪是图像处理邻域一个重要的分支。近年来随着计算机技术的高速发展，人们对图像质量的要求越来越高，图像噪声的处理已经成为了研究的一个热点。 本论文通过分析与处理有噪的数字图像，综合运用数字图像处理的理论知识以及Matlab强大的功能对加噪后的图像进行空间域、频域分析和滤波。在滤波器研究与实现的过程中，空间域滤波器采用均值滤波、中值滤波以及维纳滤波；频域滤波器采用巴特沃斯低通滤波。对比原始图像与滤波后图像，分析每种噪声的最适滤波器，得出相应的结论。通过分析与对比仿真结果得出：均值滤波器是典型的线性滤波器，对高斯噪声的抑制效果是比较好的；中值滤波器是常用的非线性滤波的方法，对椒盐噪声的滤除特别有效；维纳滤波器对高斯噪声有明显的抑制作用，但对椒盐噪声的抑制不是很理想。 关键词：图像去噪；空间域；频域分析 Abstract The noise-reduction of digital image is an important branch of Image processing fields. In recent years, people’s demand to image quality is increasingly high with the rapid development of computer technology, and the image noise processing has become an hot spot in research. This paper through the analyzing and processing of noised digital images, comprehensive using theoretical knowledge of the digital image processing and the Matlab’s powerful function in spatial domain and frequency domain analysis and filtering for the images after adding noise. In the process of filters researching and implementation, filters in spatial domain usually adopt the mean filtering, median filtering and wiener filtering, but filters in frequency domain usually adopted Butterworth low-pass filtering. According the simulation result’s analysis and comparing we know: the mean filter is a typical linear filter, the inhibition effect of Gaussian noise is better; the median filter is a common method of nonlinear filtering, the inhibition effect of salt and pepper noise is better; the wiener filter has obvious inhibitory effect to the gaussian noise, but the salt and pepper noise suppression is not very ideal. Keywords: The noise-reduction of image;Spatial doma