P-M模型扩散函数的改进

Improved Diffusion Function of P-M Model

P-M模型是图像平滑中经典的模型,其特点是既能消除孤立噪声点,又能有效保持图像边缘。P-M模型中的扩散函数,其作用是控制平滑力度。本文在研究P-M模型基础上构造了一个新的扩散函数,采用新扩散函数的P-M模型对图像进行平滑,其结果与经典的P-M模型相比,在峰值信噪比相当的情况下,迭代次数大大减少,运行速度得到很大提高。

P-M model is a classical model for image smoothing, the P-M model which is able to remove the isolated noise as well as preserve the edge effectively. The diffusion function is used for controling the smoothing degree. In this paper, a novel diffusion function is introduced based on researching P-M model. The novel diffusion function is adopted in the P-M model and experiment results show that when the Peak Siganl-to-Noise Ratio（PSNR） decreases the iterative times almost equivalent greatly and increases the process speed prodigiously, compared with the classical P-M model.