基于改进小波阈值法的图像去噪算法

The Algorithm of Image Denoising Based on the Improved Method of Wavelet Thresholding

图像去噪是图像处理中最基本、最重要的前期工作,本文提出一种基于衰减法的Garrote阈值函数,并将基于该改进阈值函数的小波阈值法用于图像去噪过程,最后通过MATLAB仿真实验验证了本文所提出算法的有效性。本文在分析小波阈值法对图像去噪效果影响的基础上,针对该去噪算法在去除噪声的同时也损失了一定量的图像细节信息的问题,改进了传统阈值函数未考虑阈值以下的小波系数可能含有图像细节信息而对阈值以下小波系数盲目置零的缺点,对Garrote阈值函数阈值以下的小波系数采取衰减方法,以保留更多的图像细节信息,并加入三个调整因子以提高其性能和灵活度,实验表明本文提出的改进小波阈值去噪算法能够有效地去除噪声,且能够保留大量的图像边缘及细节信息。

Image denoising is the most basic and important preliminary in work image processing,A threshold function of Garrote based on the attenuation law is proposed,and a wavelet thresholding method based on the improved threshold function is used in image denoising process.At last,the effectiveness of the algorithm improved by this paper is verified through the MATLAB simulation experiments.With analyzing the influences on the effect of image denoising by the method of wavelet threshold,and for the problem of a certain amount of image detail information is lost while the noise is removed by the denoising algorithm.And the shortcomings that the wavelet coefficients below the threshold is set to zero blindly without considering the wavelet coefficients below the threshold may contain the image detail information are improved.In order to retain the more image detail information,a attenuation method is taken in below a threshold value of the wavelet coefficients of Garrote threshold function.The experiments show that the noise can be effectively removed and a large number of image edges and details can be retained by the denoising algorithm based on the improved wavelet threshold.