干涉SAR图像的降噪方法及水平地形效应消除

考察不同滤波器抑制干涉SAR(Synthetic Aperture Radar)图像噪声的效果.为有效地使用低通滤波处理,首先对原始干涉图像进行水平地形效应消除,保证图像信号频率的中心在零频处,降低由干涉SAR系统几何结构引起的信号频率偏移对滤波效果的影响.同时水平地形效应消除后的干涉SAR相位图可直观地反映地形的拓扑结构.实际数据处理结果证明本文提出的水平地形效应消除算法及滤波算法的可行性及有效性.

Flat Earth Removal and Denoising in InSAR Images

Thermal and "Speckle" noise is an obstacle to generate digital elevation model(DEM) from the interferogram by 2-D unwrapping of Interferometric SAR. The boxing and median filters effects on reducing the noise are investigated in detail in this paper. Especially for the single complex images, it is more convenience to process the images by low-pass filters in image domain than in the frequency domain. And the window size effects on the processing are also discussed in detail. In order to process the noise by low pass filter reasonably, the flat earth removal is introduced to eliminate the frequency shift effect arisen from the geometry of the interferometric SAR. At the same time, the interferogram after earth flat removal can show the earth topography directly. The raw data show the validation of the algorithms in the processing.