一种改进的基于各向异性扩散方程的图像去噪方法

Improved Algorithm of Image Denoising Based on Anisotropic Diffusion Equation

在Perona-Malik非线性扩散滤波方法的基础上,提出一种新的扩散函数,该函数不需要预先设定经验参数,能够根据图像的噪声程度,在每次迭代过程中实时调整扩散参数.改进后的图像去噪方法在实验样本上进行了对比实验,结果表明,该方法不仅去噪保边缘的效果很好,而且迭代次数明显减少,提高了图像处理的效率.

Based of Perona-Malik nonlinear diffusion filtering method,this paper proposes a new diffusion function.The function does not need to set the experience parameters in advance.And according to the image noise level,it could adjustment of diffusion parameters in each iteration process.The experimental results show that this method not only has good denoising and edge preserving effect,but also reduces the number of iterations obviously and improves the efficiency of image processing.