一种提高INSAR干涉图质量的新方法

A New Method for Improving the Quality of INSAR Interferogram

如何获取高质量的干涉图是INSAR技术的核心问题之一。首先概述了干涉图中存在的噪声,并对干涉图中频谱混叠现象的成因及其影响进行了重点分析。针对SAR影像距离向和方位向频谱的特性,通过增加影像频谱中空白带的带宽对影像进行过采样,从而成功消除了干涉图中的频谱混叠现象。针对过采样后干涉图数据量激增的问题,利用干涉图频谱的理论权函数推导了欠采样函数。最后,利用ERS数据验证了文中的过采样方法能有效消除干涉图中的频谱混叠现象,显著提高干涉图质量。同时,利用文中推导的欠采样函数可以使干涉图在恢复原始大小的同时,使噪声得到进一步抑制。

One of core problems in INSAR technology is how to obtain the interferogram with high quality. This paper first summarizes the noise in interferogram, and analyses the sources and influence of spectrum aliasing in the interferogram. Based on the characteristic of range and azimuth spectrum of SAR images, oversampling is used to remove the aliasing, which means broadening the empty band in spectrum artificially. By this way the aliasing in the interferogram is eliminated successfully. Because the size of interferogram is many times larger than the original interferogram, the downsampling function is deduced. At last the oversampling and downsampling theories are tested with ERS satellite images. Test results show that the oversampling method can remove the spectrum aliasing successfully. And the downsampling method can not only restore the size of interferogram but also improve its quality remarkably.