In deformation analysis, invariant strain descriptors that are independent of the coordinate system are significant. An approach is proposed to determine the strain tensor and invariant strain quantities through the 3D retrieved co-event displacement of differential interferometric synthetic aperture radar (DInSAR). An optimal cell count is selected based on the proposed rule-of-thumb formula. This set of coherent cells with prescribed acceptable minimum value coherence is distributed on the domain through a quasi-random number generator. The recursive moving least squares method is applied to estimate the components of strain and differential rotation tensors. In practice, this method is not limited to the homogeneous strain field hypothesis. The objective here is to propose an approach to determine the local and regional strain patterns in SAR’s image resolution details. This approach is tested on both synthetic and real data. The real data consist of 3D displacements of 19 global navigation satellite system stations and SAR images of Kilauea Volcano eruption in July 2007. Results indicate that the areas with maximum shear strain and maximum dilatation are located exactly near the east rift zone of the volcano.

在变形分析中，独立于坐标系的不变应变描述符非常重要。提出了一种通过差分干涉合成孔径雷达（DInSAR）的3D检索共事件位移来确定应变张量和不变应变量的方法。基于建议的经验法则公式选择最佳的细胞数。通过准随机数发生器在域上分配具有规定的可接受的最小值相干性的这组相干单元。应用递推移动最小二乘法估计应变和旋转差张量的分量。实际上，该方法不限于均质应变场假设。这里的目的是提出一种确定SAR图像分辨率细节中局部和区域应变模式的方法。此方法已在综合数据和真实数据上进行了测试。实际数据包括19个全球导航卫星系统站的3D位移和基拉韦厄火山喷发的SAR图像（2007年7月）。结果表明，具有最大剪切应变和最大膨胀率的区域恰好位于火山东纵裂带附近。