基于PDE方法的遥感图像处理

Remote Sensing Image Preprocessing Based on PDE Method

前期处理后的遥感图像中主要剩有高斯白噪声和椒盐噪声,传统平滑滤波技术很难同时兼顾除噪和保持边缘的要求;所改进的偏微分方程(PDE)滤波算法能很好地适应遥感图像噪声特征,在消除噪声的同时能很好的保留边缘和纹理细节;仿真实验验证了算法的正确性及有效性;研究了算法关键参数的选择范围和处理效果,为实际应用提供指导。

After pre-filtering, the residual noise of remote sensing image can be regarded as Gauss noise and pepper-salt noise. It＇ s very difficult to filter these noises while to keep the edges in the image. A new improved PDE model is introduced and applied to the remote sensing image, which can deal with this problem. By chosing appropriate parameters, the results are encouraging.