基于小波域上各向异性扩散的图像去噪算法

给出一种将小波阈值和偏微分方程相结合的去噪算法。首先对图像进行小波分解,在小波域上用改进的Perona Malik模型进行各向异性扩散,保留或较小程度的伸缩幅值大的系数,平滑幅值小的系数,较好地保留自然图像在小波域上的非线性相关性的同时实现小波系数伸缩,然后进行小波重构。仿真实验表明,该算法能获得较高的信噪比和较好的主观质量,且运算量比Perona Malik方法要小。

Image Denoising Algorithm Based on Anisotropic Diffusion in Wavelet Domain

An image denoising algorithm combining wavelet threshold and PDE is proposed.The image is mapped to wavelet domain using wavelet decomposition,by anisotropic diffusing with modified PeronaMalik equation in wavelet domain,to keep coefficients of high amplitude almost untouched and smooth off the others,to shrink coefficients with much nonlinear correlation still remained in wavelet domain,then reconstruct back using inverse wavelet transform.Computer experiments indicate that the present algorithm is efficient for image denoising and shorter time consuming than PeronaMalik Model.