一种混合噪声图像的自适应滤除方法

基于小波变换的阈值去噪方法仅适用于去除高斯白噪声,对于脉冲噪声得不到好的去噪效果,正交小波变换由于缺乏平移不变性,在去噪过程中会产生人为的振荡现象,使图像边缘失真,甚至图像模糊,提出了基于平稳小波域自适应阈值算法同中值滤波相结合的去噪方法,该方法能有效地滤除图像中的高斯白噪声和脉冲噪声组成的混合噪声,并验证了该方法的有效性。

A adaptive de-noising method used to image with mixed noises

The de - noising method by threshold based on wavelet transform is only used to remove Gaussian white noise and invalid to remove impulse noise, and orthogonal wavelet transform can get result with Gauss noise, but may produce artifacts on discontinuities of signal. Which causes image edge distortion seriously. The reason is that the denoising algorithm lacks of wavelet translation invariant. This paper proposes the de - noising method that the median filter in company with the stationary wavelet adaptive threshold algorithm at the same time, this method can effectively remove fixed noises of the Gaussian white noise and impulse noise. The validity of algorithm is proved.