摘要 为解决经典PSInSAR技术在非城区因受永久散射体空间分布不足而导致地形形变监测误差较大的问题，提出基于分时散射体(partial time scatterer，PTS)提取的改进算法。首先基于改进的经验模态分解对影像进行边缘保持平滑滤波降噪，然后采用可信概率估计对PTS目标进行联合提取，最后通过参数差分估计分离PTS相位和计算形变速率，从而得到监测区的地表形变。实验结果表明，提取的PTS目标基本可保持传统PS点的空间分布特性和时序变化趋势，提高非城区目标点的空间分布密度，本文算法具有有效性。
Abstract：To solve the problem of large monitoring error caused by insufficient spatial distribution of permanent scatterers when PSInSAR is applied to non-urban surface deformation monitoring, we propose an improved PSInSAR algorithm based on partial time scatterers(PTS). Firstly, the image is denoised by edge preserving filtering based on the improved empirical mode decomposition. Then, the PTS target is extracted by the credible probability estimation. Finally, the PTS phase is separated by parameter difference estimation and the deformation rate is calculated to obtain the surface deformation of the monitoring area. The experimental results show that the extracted PTS target basically keeps the spatial distribution characteristics and temporal variation trend of traditional PS points, which improves the spatial distribution density of non-urban target points and verifies the effectiveness of the algorithm.
