实现自适应边缘增强的P-M模型

P-M模型利用扩散偏微分方程进行图像平滑，通过随梯度自适应变化的扩散系数实现去除噪声的同时保护图像边缘特征，为进一步提高图像视觉效果，希望在图像边缘附近进行逆扩散以增强边缘特征，为此提出了实现自适应边缘增强的P-M模型。该改进模型中的边缘增强作用基于shock滤波器逆扩散方程，通过设置随梯度自适应变化的逆扩散系数而实现。实验结果表明，相比经典P-M模型，改进模型能使去噪后的图像有更好的主观视觉效果，同时峰值信噪比也更高。

Adaptive edge enhancement P-M model

P-M model is an image smoothing method in which diffusion PDEs are used to realize not only image denoising but also preserving image edge character by the diffusion coefficient adjusted with image grads.For furthermore improving the image visual effect,the inverse diffusion is wished to make nearby the image edge,in order to enhance the edge character.Therefore,an adaptive edge enhancement P-M model is proposed.In this modified P-M model,edge enhancement based on shock filter is realized by setting the inverse diffusion coefficient also adjusted with image grads.The experiment results show that compared with the traditional P-M model,the modified P-M model can get not only better visual effect but also bigger PSNR.