一种结合最优相干运算的极化干涉SAR相干配准方法

针对丰富植被地区散射中心高度差去相关的问题,提出了一种结合最优相干运算的Pol-In-SAR相干配准方法,讨论了Pol-InSAR复图像配准过程中需要考虑的几个关键问题,并以天山地区SIR-C/SLC复图像的配准结果证明了方法的有效性。

Registration of Pol-InSAR Images Combined with Coherence Optimization

Polarimetric synthetic aperture radar (SAR) interferometry, which combines SAR polarimetry with SAR interferometry, can improve the measurement precision of interferometric SAR and enhance the ability to explain the scattering mechanisms of targets. So it has become an important direction of SAR application development. In the area coverd by rich vegetations, the confusion of scatting centers induces considerable loss of coherence, which is an important element of interferometry and tightly related with the quality of interferometry. To reduce the loss of coherence, polarimetric information was induced in SAR interferometry because of its potential ability of separating scatting centers. Traditional Pol-InSAR flow worked on this problem during the generation of interferometric phase maps. It ignored a question that the loss of coherence can influence the registration as well because the coherence is one of estimate values for registration. As we all know, registration is an important prepositive key step, its quality is crucial for interferometry. All bad factors of registration should be reduced. Considering above, a new method for registration of Pol-InSAR images combined with coherence optimization is brought forward. The coherence optimization can separate scatting centers, it will availably reduced the loss of coherence induced by confusion of scatting centers. An experiment on SIR-C/SLC complex images demonstrates the validity of this method.