提出一种改进的永久散射体PS (Persistent Scatterer)探测方法,首先基于时序SAR影像的振幅信息,利用振幅和振幅离差指数双阈值法获取PS候选点,然后根据干涉相位空间相关性,对所有候选点进行相位稳定性分析,筛选出既满足雷达反射强度高、又符合散射稳定性要求的真实PS点,最后基于所选PS点提取形变信息。选取上海市局部区域作为实验区,以16幅TerraSAR-X (TSX)影像为数据源,进行PS探测和形变信息提取。结果表明,相比于已有的PS探测方法,本文方法能够更准确、可靠地识别出时序SAR影像中的PS点。形变结果显示,研究区域内最大沉降速率为-51 mm/a,平均沉降速率为-15.8 mm/a,以水准实测数据进行检核,验证了基于改进的PS探测方法进行地表形变信息提取结果的可靠性。

An improved method for Persistent Scatterer (PS) detection was presented in this paper. Based on the time series of SAR amplitude data, the PS Candidates (PSCs) were first selected using both amplitude threshold and amplitude dispersion index threshold. The true PSs were then screened out from the PS Candidate (PSC) set by an iterative process based on phase stability analysis. The deformation values of the true PSs were finally estimated. The experiments of PS detection and deformation extraction were performed using 16 high resolution SAR images collected by the X-band radar sensor onboard the German satellite TerraSAR-X covering Shanghai. The test results demonstrated that the PS points detected by the improved method were more accurate and reliable than those identified by traditional methods. The maximum subsidence rate in the investigated area reached -51 mm/a and the average subsidence rate was -15.8 mm/a. The subsidence rate error derived by PSI was ±4.45 mm/a compared with ground-based measurements derived by precise leveling. This confirms that the PSI technique is reliable for extracting regional ground subsidence using the proposed method.