InSAR图像相位解缠的最小费用流法及其改进算法研究

Minimum cost flow and its ameliorative algorithm for phase unwrapping of InSAR image

最小费用流法是基于网络流的相位解缠方法,解决了许多解缠方法无法消除相位噪声对高相干区域影响的问题,在此基础上,本文针对该方法解缠时速度较慢和对计算机性能要求较高的缺点而提出改进算法,即将干涉图像分为若干子区域分别进行处理,再利用基于Contourlet变换的超小波方法进行融合处理,最后用算例进行了验证,结果表明最小费用流法及其改进算法是一个较好的解缠方法。

Minimum Cost Flow （MCF） based on network flow is a phase unwrapping algorithm which can eliminate influence of phase noise for the more correlative district while many other algorithms cannot do. Based on MCF, ameliorative MCF was put forward aiming at disadvantages for example low speed and demanding higher computer capability during phase unwrapping. Firstly, interferograms were divided into a few sub-areas so that they could be unwrapped respectively. Secondly, interferograms were fused by supper wavelet based on contourlet transform. In the end, Minimum Cost Flow and its ameliorative algorithm were validated by practical example, and results indicated that the method was effective in phase unwrapping.