维纳滤波和非线性扩散相结合的图像去噪

提出一种基于小波和非线性扩散的新的图像去噪算法。小波域局部维纳滤波是一种简单有效的去噪方法,利用该方法先对原始图像进行初步去噪,以此引导非线性扩散模型中的边缘检测函数,再用非线性扩散进行去噪。实验表明:该算法不仅很好地保存了图像的边缘信息,而且有效地去除了图像中的大部分噪声,无论是视觉效果还是客观标准上都优于单纯的小波域维纳滤波或非线性扩散去噪。

Image Denoising Based on Wiener Filtering in Wavelet Domain and Nonlinear Diffusion

A new image denoising method based on wavelet and nonlinear diffusion is presented.Among the popular wavelet-based image denoising methods,locally Wiener filtering with thresholding is a simple but efficient one,which is applied to obtain a cleaner image.Then this cleaner version is used as a guide to get the edge testing function in nonlinear diffusion to reduce noise in the original image.Experimental results show that important features in the image,such as edges,can be well recovered and most noise components in uniform parts can be removed.The new method performs well in view of both the common used objective standards and the visual effect.