InSAR相位解缠算法的分析与评价

Analysis and evaluation of InSAR phase unwrapping algorithms

相位解缠是合成孔径雷达干涉测量中的一个关键步骤,是主要的误差源之一,同时也是研究InSAR的一个重点和难点。相位解缠算法的选取对InSAR最终的成果(DEM或者形变量)有很大的影响。文中选取常用的Gold-stein枝切算法、预解共轭梯度法(PCG)以及最为流行的统计费用网络流算法(Snaphu),对西安地区的实例数据进行解缠运算,采用不同的评价标准对解缠结果的质量进行评价比较,并给出各自解缠算法的特点和适用情况,最后对相位解缠算法的未来发展方向进行展望。

Phase unwrapping is crucial in InSAR and a major source of InSAR errors. At the same time it is also very important and difficult in studying InSAR technique. The result of InSAR（DEM or deformation） is different when choosing different phase unwrapping algorithms. The results of Goldstein branch-cut algorithm, Preconditiofied Conjugate Gradient algorithm and Snaphu algorithm in Xi＇an are compared. The unwrapping results is evaluated by several criteria. The characteristic and application scope of each algorithm is analyzed, and the prospect of phase unwrapping algorithms is discussed.