测绘学报

• 学术论文 •    

基于边界探测的InSAR最小二乘整周相位解缠方法

陈强,杨莹辉,刘伟堂   

  1. 西南交通大学
  • 收稿日期:2011-03-24 修回日期:2011-10-11 出版日期:2012-06-25 发布日期:2012-06-25
  • 通讯作者: 陈强

InSAR Phase Unwrapping Using Least Squares Method with Integer Ambiguity Resolution and Edge Detection

  • Received:2011-03-24 Revised:2011-10-11 Online:2012-06-25 Published:2012-06-25

摘要: 为解决最小二乘相位解缠方法一直存在收敛性差和非整周性相位解算问题,本文提出基于解缠边界探测并搜索整周相位值的解缠方法,采用最小二乘次优解与最优解的阈值判别准则,预测解缠趋势并逐步推进相位解缠边界,通过在解缠迭代计算中附加相位搜索增量以提取相位整周值。实验分别基于噪声干扰的模拟图与ERS-1/2卫星覆盖香港和深圳地区的SAR影像干涉图,采用边界探测算法开展相位解缠计算,试验结果表明,本文解缠方法对附加强噪声的模拟干涉图能正确解缠出99.5%的相位值,真实SAR影像干涉图的解缠精度达到±0.12rad。

Abstract: In order to overcome the problems of poor convergence and integer ambiguity solution in phase unwrapping based on least squares method, an improved method of phase unwrapping is proposed based on unwrapping edge detection in this paper. The unwrapping edge is predicted and propelled by comparing the optimal solutions and suboptimal solution in least squares method. The integer ambiguity of unwrapped phase is detected and extracted through an attempt to use additional phase increment into wrapped phase. The feasibility and accuracy of proposed method is tested and validated by using simulated data and ERS-1/2 InSAR interferogram covering Hong Kong and Shen Zhen area. The testing results show that the proposed method can correctly extract 99.5% phase values from the simulated interferogram with strong noise. The accuracy of phase unwrapping tested on the InSAR interferogram reaches ±0.12rad.