改进CLMC模型应用于医学图像滤波的研究

在分析CLMC模型（正则化P-M模型）的基础上，在扩散系数中加入二阶偏导数，提出了一种新的偏微分方程（PDE）模型，并采用基于半隐式方案的加法算子分裂算法（AOS）对新模型进行解算。选用MEDPIX医学图像数据库作为实验数据来源对新模型的滤波效果进行验证分析，实验结果表明，相较于CLMC模型，新模型依然能很好的平滑噪声，且在保持边缘细节方面有一定的提升。

Improved CLMC Model Used in Medical Image Filtering Research

Based on the analysis of the CLMC model(regularized P-M model) applied in image filtering, a new partial differential equation(PDE) model was presented by adding second derivative to diffusion coefficient, and it was solved by a numerical method using additional operator splitting algorithm(AOS) based on semi-implicit scheme. The paper chose medical images from MEDPIX as the experimental objects. Experimental results indicate that, comparing with CLMC, the new model shows superior performance on preserving image edge and nearly equal good denoising effects.