一种新的基于小波变换的图像去噪方法

根据图像小波分解的特点和小波分解后高频小波系数的统计特性,构造了一种新阈值函数的去噪算法。与传统的硬阈值、软阈值函数相比,新阈值函数考虑了图像能量分布的特点,对于每一小波系数乘以一个与自身大小相关的降噪因子,并且新阈值函数简单易于计算,具有较强的自适应性。实验结果表明,采用新阈值函数的去噪结果能够有效地抑制图像的马赛克效应,无论在视觉效果上,还是在信噪比增益上均优于传统的软、硬阈值方法。

A new method of image denoising based on wavelet transform

According to analysis of the characteristic of wavelet transform and statistical character of wavelet coefficient,a new wavelet threshold function was presented.Comparing with conventional soft and hard threshold function,this new threshold function has considered the distribution of image energy.Each wavelet coefficient multiplication by different denoising factor,so it is a simple and adaptive denoising algorithm.Experimental results show that the denoising algorithm can mitigate the mosaic appearance effectively and the numerical results also show that the new algorithm give better PSNR gains than conventional soft and hard threshold algorithm.