Abstract: To solve the problem of large monitoring errors caused by insufficient distribution of permanent scatterers (PS), when PS-InSAR is applied for non-urban ground deformation monitoring, an improved PS-InSAR algorithm based on partial time scatterers is proposed. Based on the analysis of a large number of partial time scatterers in non-urban areas, the partial time scatterers are extracted through the combination of dual threshold and credible probability after the image is denoised by improved EMD algorithm. And then, the deformation rate is calculated through the iterative estimation of the different parameters and the phase separation of the partial time scatterer, obtaining the surface deformation of the monitoring area. Experimental results show that the extracted partial time scatterers basically cove the PS points extracted by the traditional PS-InSAR algorithm and maintains their spatial distribution characteristics and the timing change trend. The spatial distribution density required for surface deformation monitoring in non-urban areas is greatly improved. Improved the accuracy of non-urban surface deformation monitoring, which verifies the effectiveness of the algorithm.

Key words: non-urban surface deformation monitoring, improved PS-InSAR, partial time scatterers, noise reduction of improved EMD decomposition, credible probability target extraction