优化的区域增长InSAR相位解缠算法

Optimized Region-growing Algorithm for InSAR Phase Unwrapping

在区域增长的干涉合成孔径雷达(InSAR)相位解缠算法中,种子的选择和种子相位值的确定是关键问题之一。本文研究了优化的区域增长2维相位解缠方法,包括合适的种子选取和种子相位值的优化处理,提出了将干涉相位图中相位跳变导致的边缘曲线作为种子,通过分析边缘曲线之间的相邻关系,基于遗传算法优化种子的相位值。仿真结果表明,该算法的计算量与干涉相位图的像素点数成近似线性关系,同时由于该算法属于局部算法,可以通过并行处理进一步提高计算效率。优化的区域增长2维相位解缠算法与其他2维相位解缠算法相比较,其解缠后的相位条纹与原始干涉相位图相位条纹的一致性非常好。该算法的处理结果与最小成本网络流相位解缠法相当,但其计算量却远远小于最小成本网络流相位解缠法的计算量。

For region-growing phase unwrapping algorithm of interferometric synthetic aperture radar （InSAR）, one of the key problems is to determine the seeds and the phase values of these seeds. In this paper, the optimized region-growing algorithm for 2-D phase unwrapping is discussed, involved determine the seeds and optimize the phase values of seeds. Taking the edge curves of interferogram as the seeds and analyzing the neighboring relationship of edge curves, the optimized phase values of seeds can be obtained by using the genetic algorithm. The experimental results show the computational complexity of the optimized region-growing algorithm is linearly correlative with the number of interferogram pixel. As a regional algorithm, the efficiency of the optimized region-growing algorithm can be improved by the paralleled processing. Compared with the branch-cut algorithm, the weighted least-squares algorithm and the network flow algorithm, the optimized region-growing algorithm has excellent coherence of fringe between the result image and original interferogram, as the network flow algorithm. But the computational complexity of this algorithm is far less than the network flow algorithm＇s.