1、 I 基于小波变换的图像处理方法研究 摘 要 图像增强是图像处理的一个重要分支,它对提高图像的质量起着重要的作 用。它通过有选择地强调图像中某些信息而抑制另一些信息,以改善图像的视觉 效果,将图像转换成一种更适合于人眼观察和计算机进行分析处理的形式。传统 的方法在增强图像对比度的同时也会增强图像噪声, 而小波变换是多尺度多分辨 率的分解方式,可以将噪声和信号在不同尺度上分开,根据噪声分布的规律就可 以达到图像增强的目的。 本文首先对传统图像增强理论进行概述, 并给出直方图均衡化与灰度变换算 法,通过 matlab 来观察其处理效果的特点,然后提出四种基于小波变换的图像 增强方法,并分析它们与传
2、统图像增强方法相比的优缺点,最后基于传统小波变 换只能增强图像边缘部分而无法增强细节部分的缺点, 引出了基于分数阶微分和 小波分解的图像增强方法,并通过 matlab 观察了这种算法的处理效果。 关键词:图像增强; 直方图均衡化; 小波变换; 分数阶微分 II Image enhancement based on wavelet transformation Abstract Image enhancement is an important branch in image processing.It plays an important role in improving the quali
3、ty of the images.It will improve the image visual effect through emphasizing the image information and inhibitting some other information selectively.It will converse images into a form more suitable for the human eye observation and computer analysis processing.The traditional method of image enhan
4、cement will enhance image contrast,image noise as well,while wavelet transform is a decompositon method of multi-scale and multi-resolution,it can separet noise from signal in different scale so that it can arrive the purpose of image enhancement according to the distribution of the noise. In the paper,firstly, I will summarize the image enhancement theory and give the Histogram equalization algorithm,at the same time,I will analyze the disadvantages of the treatment effect through the Matlab.Th