利用线性相位模型提高干涉SAR图像相干性及改进相干性估计

本文提出一种对干涉合成孔径雷达(干涉SAR)处理中单视复图像(SLC)进行地形自适应预滤波的方法,提高从不同入射角获得的SAR图像之间的空间相干性,并同时改进相干性估计.通过Chirp-Z变换进行干涉图局部频率估计,从而提取干涉图线性相位模型,并利用其进行距离和方位向的地形自适应预滤波.由于滤波器参数随地形进行调整,其性能优于仅仅依赖轨道参数或平坦地球相位估计的固定带宽非自适应滤波器.文中对SIR-C/X-SAR的Etna火山干涉数据同时进行了距离和方位向地形自适应处理,利用相干系数直方图对该方法与固定带宽非自适应滤波方法的性能作了比较.最后,本文还应用线性相位模型改进相干性估计,得到了去除地形变化因素后的相干系数图.

Improving the Coherence for InSAR Processing and Coherence Estimation Using the Linear Phase Model

A topography adaptive filter is proposed to improve the coherence of single look complex(SLC) images obtained from different looking angles for interferometric synthetic aperture radar(InSAR) processing,and to improve the coherence estimation as well.The extraction of the linear phase model of the interferogram via local frequency estimation based on the 2-D Chirp-Z transform makes this filter topographically adaptive both in range and azimuth.Adjusting the filter parameters according to the local topography,the adaptive filter performs better than the fixed bandwidth non-adaptive one designed through flat earth phase estimation.Interferometric data set of Mt.Etna is processed topographicaeey adaptively in range and azimuth,and the results are employed for the comparison of the performances of the proposed and conventional filters.Finally,the linear phase model is also used to improve the coherence estimation by eliminating the topography variation in the estimator.