小波分解得到的低频近似图像其实也是一个小波降噪的应用,基于边缘检测的降噪方法是它的改进。边缘是图像的高频特征,一般常用的边缘检测方法并不能很好的得到所有有用的边缘信息。
摘要:针对传统信号处理方法的不足,小波分析作为改进的理论方法在20世纪80年代被提出来。小波函数是一类函数的统称,它们是用来产生多尺度窗函数的原型,为研究信号处理提供了高效的分析工具。小波的多尺度分析方法在数值计算、信号处理、地质勘探、机械工程等很多领域有着广泛应用。然而,在处理具有各向异性奇异性的对象时,比如图像压缩,降噪等应用中,小波分析会不可避免地在图像的边缘位置造成一定程度的模糊。因此,在应用中预先提取并保护图像边缘特征是有必要的。 本文首先结合傅立叶变换阐述了小波分析的一些理论和重要性质,包括连续和离散两种变换。其次介绍了在数字图像降噪方面的一些应用,重点介绍了小波变换的阈值降噪方法。最后,本文使用Sobel算子执行图像的边缘检测,保护了图像的边缘特征,通过进行二维离散小波变换得到小波系数,执行阈值降噪的处理,实现了图像的降噪,改善了图像质量。
关键词:小波变换;图像降噪;多分辨分析;子带编码
The Application Design and Implement of Image De-nosing Based on Wavelet Analysis Theory
Abstract: The analysis of wavelet as an improved theoretical method was proposed in the 1980s according to the shortcomings of traditional signal processing methods. The wavelet function is a general term of a kind of functions, which are prototypes for generating multi-scale window functions and can provide efficient analysis tools for signal processing. Wavelet multiscale analysis is widely used in many fields such as numerical calculation, signal processing, geological exploration, mechanical engineering and so on. However, the wavelet analysis will inevitably cause a certain degree of blurring in the image edge position when dealing with some objects of anisotropic singular such as image compression, noise reduction and other applications. Therefore, it is necessary to pre-extract and protect image edge features during the application. Firstly, combining the Fourier transforms, this paper describes some theories and important properties of the wavelet analysis, including continuous and discrete transforms. Secondly, some applications about the noise reduction of digital image are introduced, especially the noise reduction of the threshold of method of wavelet transform. In the end, the Sobel operator is used to perform image edge detection, which protects image edge features. Wavelet coefficients are obtained by two-dimensional discrete wavelet transforms, and the noise reduction of threshold is performed to achieve the noise reduction of image and improve image quality.
Keywords: Wavelet Transform,Image De-nosing,Multiresolution Analysis,Subband Coding
目 录
摘要 3
关键词 3
一 前言 3
1.1小波分析概述 3
二 小波分析产生背景 4
2.1傅立叶变换(FT) 4
2.2短时傅立叶变换(STFT) 5
三 小波分析理论阐述 5
3.1连续小波变换(CWT)的定义 5
3.2 小波逆变换 6
3.3 离散小波变换(DWT) 6
四 数字图像处理概述 8
4.1 数字图像 8
4.2 图像处理 8
4.3 MATLAB支持的基本图像类型 9
五 小波降噪应用与实现 9
5.1 图像的二维小波变换 9
5.2 图像的边缘检测 11
5.3小波图像降噪方法简述 13
5.4阈值的选定及阈值处理方法 13
5.5基于边缘检测的小波图像阈值降噪方法 14
六 结束语 16
致谢 17
参考文献 17
基于小波分析的图像降噪应用设计与实现
一 前言
1.1小波分析概述
在30多年之前,由于信号处理的实际需求,小波分析开始受到关注,它的理论和研究得到发展。在信号处理中最常用到的就是傅立叶变换,然而傅立叶变换不能在时域和频域同一时间提供信号的小区域细节信息,相比之下小波变换在这个问题上有更好的表现,所以可以更有效率的从处理后的信号中提取出有用的信息。