基于二进小波变换的阈值图像去噪方法研究

发布时间：2018-03-01 10:13

本文关键词： 图像去噪高阶消失矩二进小波滤波器阈值函数　出处：《新疆师范大学》2017年硕士论文　论文类型：学位论文

【摘要】：现实生活中,噪声的存在依旧是无法忽视的问题。近些年来,随着科技的不断发展,及小波理论的进一步的填充,小波在图像去噪中变得越来越有优势,其已逐步成为图像处理的重要研究内容之一。小波变换领域中正交小波研究居多,但是其下采样,使得图像变换到小波域后会造成图像位置信息发生偏差,从而引起视觉上的变形。故文章构造出具有高阶消失矩的二进小波滤波器和卷积型正交小波,还提出一种新的阈值函数同时作用对图像进行去噪处理,并将去噪后的图象的PSNR数据和效果进行比较与分析。最终由仿真实验得出,基于二进小波的阈值去噪方法是一种有效的图像去噪方法。主要研究内容如下:(1)介绍本课题的研究背景和意义,并将小波阈值去噪的发展和现状、文章的结构与安排、文章的创新点进行简述。(2)小波变换的基本的理论,简单的叙述连续、离散小波变换的定义和相关的定理、重点去介绍与二进小波相关的一些理论,最后将详细介绍Mallat算法和a?trous算法这两种快速算法,在一定程度上突出这两种算法的差异。(3)在深入分析软、硬阈值函数的优缺点的基础上,引入改进的阈值函数,并构造出卷积型正交小波对图像进行去噪分析;(4)正交小波变换不具有平移不变性,使得去噪后的图像容易出现Gibbs现象而不能满足人类视觉要求。为了抑制这种现象,引入了具有平移不变性的二进小波变换,构造具有高阶消失矩的二进小波滤波器结合改进的阈值函数,提出基于二进小波变换的阈值去噪方法。仿真实验结果表明,该方法无论在峰值信噪比方面还是在视觉效果方面与小波硬阈值、软阈值去噪方法相比都有明显的提高和改善,是一种有效的图像去噪方法。
[Abstract]:In real life, the existence of noise is still a problem that can not be ignored. In recent years, with the development of science and technology and the further filling of wavelet theory, wavelet has become more and more advantageous in image denoising. It has gradually become one of the important research contents of image processing. In the field of wavelet transform, the research on orthogonal wavelet is more and more, but its down-sampling results in the deviation of image position information when the image is transformed into wavelet domain. Therefore, a dyadic wavelet filter with high order vanishing moments and convolutional orthogonal wavelet are constructed, and a new threshold function is proposed to de-noise the image simultaneously. The PSNR data and effect of the denoised image are compared and analyzed. The threshold denoising method based on binary wavelet is an effective image denoising method. The main research contents are as follows: 1) introduce the research background and significance of this subject, and introduce the development and present situation of wavelet threshold denoising, the structure and arrangement of this paper. The innovation of this paper is to introduce the basic theory of wavelet transform, simply describe the definition of continuous wavelet transform, discrete wavelet transform and related theorems, focus on introducing some theories related to dyadic wavelet, and finally introduce Mallat algorithm and a? On the basis of analyzing the advantages and disadvantages of the soft and hard threshold functions, the improved threshold function is introduced. A convolutional orthogonal wavelet is constructed for denoising the image. (4) the orthogonal wavelet transform has no translation invariance, which makes the denoised image prone to Gibbs phenomenon and unable to meet the human visual requirements. In this paper, a binary wavelet transform with translational invariance is introduced to construct a dyadic wavelet filter with high order vanishing moments combined with an improved threshold function. A threshold denoising method based on dyadic wavelet transform is proposed. The simulation results show that, Compared with wavelet hard threshold and soft threshold denoising method, this method is an effective image denoising method, both in the aspect of peak signal-to-noise ratio (PSNR) and visual effect.
【学位授予单位】：新疆师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】