Abstract: In the process of image denoising, in order to remove noise effectively and preserve edges and key details, the diffusion coefficient based on the Cattle model is improved and a more effective adaptive denoising model is proposed. The model can not only adopt different diffusion coefficient according to different sizes of the gradient but also lead the edge sharping factor of second order partial deviation into the diffusion coefficient. The best stop time evaluation criteria based on correlation coefficient is proposed in the mean time. The experimental results show that the improved model is superior to C model, and can better coincide with the judge standard.

Key words: image denoising, anisotropic diffusion, diffusion coefficient, correlation coefficient

摘要： 图像去噪过程中，为了在有效平滑噪声的同时较好地保护图像的边缘和细节，在Cattle平滑模型基础上，对扩散系数作出改进，提出了更有效的自适应去噪模型。该模型不仅针对不同的梯度大小采用了不同的扩散系数，而且将边缘锐化因子二阶偏导引入到扩散系数中。而在图像质量评判标准中，提出了基于相关系数函数的最佳停止时间评判准则。实验结果表明，改进的模型优于C模型，且能更好地吻合评判准则。

关键词: 图像去噪, 各向异性扩散, 扩散系数, 相关系数