基于Navier-Stokes方程的干涉相位图处理

Filtering Interferometric Phase Image Based on Navier-Stokes Equation

干涉合成孔径声纳(INSAS)形成的干涉图中,条纹代表着地形高度信息。要对干涉相位图中由系统和环境引起的很多散斑噪声进行滤波,同时,保留边缘形成的结构和不连续性。文章考虑用一个新的偏微分方程(PDE)模型去对基于Perona-Malik(P-M)方程和流体力学中的Navier-Stokes(N-S)方程的干涉图进行滤波。实验结果表明,该方法能够有效地对相位图像进行去噪并且可以保护图像边缘结构。

Interferogram from the Interferometric Synthetic Aperture Sonar（INSAS）has fringes representing the terrain height information.The interferometric phase images with much speckle noise caused by the system and environment are filtered,at the same time the structures and discontinuities formed by the fringes are preserved.In this paper,a new partial differential equation（PDE）method is considered to filter the phase images based on the Perona-Malik（P-M）equation and Navier-Stokes（N-S）equations for fluid dynamics.Experimental results indicate that this method is able to de-noise the phase images and preserve the image fringes structure effectively.