针对其他路径跟踪相位解缠算法对密集残差点的干涉图像解缠的不足,提出适用于残差点不同分布的两种路径跟踪相位解缠算法:区域分片连接法和中心扩展搜索法.这两种方法能有效地避免过长的枝切线和残差点的多次连接,既提高了解缠精度,又保留了路径跟踪法在运算速度上的优势.将上述方法应用于ENVISAT-ASAR图像,结果证明了本文方法是一种有效的InSAR处理方法.

This paper introduces a procedure of Interferometric SAR ( InSAR) processing, including the algorithms of image registration and phase unwrapping. A coarse-to-fine fast registration framework, based on themean square difference function, is proposed. Aiming at the interferogram with large size and dense residues, two unwrapping algorithms arepresented: one is themulti- patchesmethod to connectopposite residues, and another one the center expanding method to search opposite residues. These two algorithms can avoid long branch cuts and excessive connection of residues, and arewellemployed in the experiments ofENVISAT- ASAR datasets. Comparedwith conventionalmethods, our results of registration and unwrapping show good advantages in InSAR data processing.