You-Kaveh图像去噪模型扩散系数的改进

本文在You-Kaveh模型的基础上,提出了一个新的扩散系数,得到了一个去噪效果更好的方程,新方程不但能够去除高斯噪声,而且能够很好地去除椒盐噪声。同时,改进了模型中拉普拉斯算子的离散形式,使其包含更多的图像信息,能够更准确地判断图像的特征。采用本文方法处理后的图像,避免了用二阶偏微分方程处理图像常出现的"阶梯"效应;同时,和同类的四阶偏微分方程去噪模型相比,本文方法的处理结果不会出现"斑"点,因此视觉效果更加理想。最后通过实验证明了该方法的有效性。

An Improved Coefficient of the You-Kaveh's Image Denoising Model

In this paper,a new diffusion coefficient is proposed,and a new partial differential equation denoising model is obtained. The new model can not only remove the Gaussian noise,but also remove the salt-pepper noise. Meanwhile,the discretization scheme of the Laplace operator is improved,and more image information can be contained. Thus,the image features can be judged more accurately. The images processed by the new model do not incur the blocky effect which is widely seen in the images processed by the seco...