基于小波变换的图像混合噪声自适应滤除算法

Adaptive filtering algorithm for mixed noise image based on wavelet transform

为同时滤除图像中的椒盐噪声和高斯噪声,提出了一种基于小波变换的混合噪声自适应滤除算法,该算法首先采用中值滤波去除椒盐噪声,然后借助边缘检测算子区将图像为分边缘与非边缘区域,进一步对非边缘区域引入改进的均值滤波器,有效削弱高斯噪声的同时保护图像边缘细节,既初步削弱高斯噪声又保护了边缘,最后采用改进的小波阈值滤波算法,对不同的小波系数采用不同的阈值函数,通过线性回归得到各最优阈值关系式。实验结果表明,该混合噪声自适应滤除算法能有效滤除椒盐噪声和高斯噪声,在图像主观质量和客观质量上均取得了较好的效果,能提高去噪图像峰值信噪比0.5~2.0 dB。

To filter out salt and pepper noise and Gaussian noise,this paper presents a wavelet-based adaptive filtering algorithm for mixed noise.The median filter is firstly applied to remove salt and pepper noise and the edge detection is used to distinguish between the non-edge region and the edge region.Then an improved average filter is adopted on the non-edge region,initially weakening the Gaussian noise,without blurring the edge.Finally,an improved wavelet threshold filtering algorithm is utilized on the entire image,using different threshold functions on different sub-bands.The parameters of these threshold functions are gained in a linear regression process.The experimental results show that the proposed adaptive mixed-noise filtering algorithm can effectively remove salt and pepper noise and Gaussian noise.Both the subjective and objective quality of the denoising result is superior,compared with other algorithms.An increase of peak signal-to-noise ratio about 0.5 to 2.0 dB could be achieved by the proposed algorithm.