提出了一种基于新型扩散系数函数的图像去噪方法,这种扩散系数函数是利用图像特征的.一些特殊性质构造出来的。该函数较常用的P-M启发性扩散系数函数,在图像去噪的过程中,不仅能更好地抑制、平滑噪声,较完整保留了图像的细节,使图像更平滑清晰,而且还大大降低了参数的选取难度以及参数对实验效果的影响程度。并通过基于该方法的四阶非线性扩散的图像去噪实验,结果表明了该方法的有效性。
A new method of image de-noising based on a novel diffusion coefficient function is proposed, which is constructed by using some special characteristics of image features.Compared with the well-known P-M heuristic function of diffusion coefcient, this function can not ouly suppress and smooth the noise better,retain as more image details as possible,make the image more smooth and clear,but also can greatly reduce the difficulty of parameter selection and the degree of influence of parameter to experimental results.The effectiveness of this method can be proved by using the fourth-order nonlinear diffusion of/mage de-noising experiment which is based on this method.