基于小波变换的数字图像降噪处理.docVIP

基于小波变换的数字图像降噪处理 摘要：图像一般都会受到噪声污染的影响，由于噪声影响图像的输入、输出等环节，使得图像分辨率下降，同时破坏了图像的精细结构，给图像的后续处理带来不便，因此如何有效地抑制噪声已成为图像处理的首要任务。图像降噪是为了提高图像的信噪比，突出图像的期望特征。传统的滤除噪声的方法难题就是如何兼顾降低图像噪声和保留图像细节两个方面。将图像的高频成分滤掉的低通滤波方法，虽然可以有效地滤除噪声．但是高频分量衰减得越多，图像就越模糊。而小波变换技术具有可进行时域和频域的局部分析、灵活地对信号局部的奇异特征进行提取和时变滤波等特点[1]，优于傅立叶变换，能有效地处理信号，把真实的信号和噪声区分开来，所以它被普遍用于信号与图像处理。 本文结合了MATLAB进行了小波阈值图像处理，MATLAB是一种高效的工程计算语言，在数值计算、数据处理、图像处理、神经网络、小波分析等方面都有广泛的应用。同时本文既概述了小波阈值去噪原理也对图像处理后的结果进行了定性分析。 关键词：噪声污染 信噪比 滤波 小波变换 MATALAB Abstract Images are generally affected by noise pollution, because of the influence of noise of image input, output, makes the image resolution to drop, at the same time destroyed the image of the fine structure, causing inconvenience to the subsequent processing of the image, so how to effectively suppress noise has become the primary task of image processing. Image noise reduction is in order to improve the SNR of image, highlight the desired features of the image. The traditional method of noise filtering problem is how to reduce the image noise and image details two aspects. The image of the high frequency component filter out low pass filtering method, although can effectively filter out noise. But the more high frequency attenuation, image is blurred. And wavelet transform technology for local analysis of time domain and frequency domain, the flexibility to singular characteristics of local signal extraction and time-varying filtering etc[1] . is superior to the Fourier transform, the signal can be effectively dealt with, to distinguish the real signal and noise, so it is widely used in signal and image processing. This paper combines MATLAB wavelet threshold image processing, MATLAB language is a kind of efficient engineering calculation, the numerical calculation, data processing, image processing, neural network and wavelet analysis has a wide range of applications. At the same time, this article summarizes the wavelet threshold denoising both principle a