Research On GB-InSAR Microdeformation Detection Method

Micro-deformation is a common phenomenon in nature,which is usually accompanied by changes in the structure,density and compactness of objects,resulting in geological disasters and building collapse events.To a certain extent,the disasters caused by small deformation can damage people’s lives or property.Therefore,the monitoring of small deformation of objects has become an urgent demand in the current industry.Compared with most traditional deformation monitoring technologies,ground-based synthetic Aperture Radar(GB-SAR)technology has the characteristics of high measurement accuracy,long measurement distance,all-weather and all-sky time.Therefore,This technique has become a research hotspot of microdeformation monitoring.In this thesis,the GB-SAR system based on frequency modelled continuous wave(FMCW)is studied to explore its deformation monitoring principle.By improving the imaging algorithm and deformation monitoring algorithm,the monitoring ability of GB-SAR system based on FMCW for small deformation is further improved.The main contents and achievements of this thesis are as follows:(1)Based on the principle of ground-based synthetic aperture radar monitoring small deformation,the GB-SAR system model is established in line with the actual demand.By combining frequency modulated continuous wave technology and interferometry technology,the GB-SAR micro-deformation monitoring system is successfully built.A simulation program is designed to simulate the micro-deformation of GB-SAR measurement,and the rationality of system parameters and the correctness of deformation estimation method are verified.Finally,combined with hardware characteristics and multi-threading technology,a GB-SAR software system which can effectively monitor the small deformation of the target is designed.(2)To meet the requirements of GB-SAR deformation monitoring system for real-time and high-precision processing results,this thesis proposes an improved classification algorithm based on K-Means++ combined with amplitude deviation and phase variance by analyzing the traditional Permannent Scatterer(PS)selection method.This method can efficiently screen PS points with high SNR in SAR images.The whole processing process of the improved method can be roughly divided into two parts.Firstly,the amplitude deviation information set of each GB-SAR images is taken as the input of K-Means++algorithm,and the information set is divided into two categories by clustering algorithm,among which the pixels with low average amplitude deviation are divided into candidate PS points.Secondly,the phase variance information set of candidate PS points is used as the input of k-means++ algorithm,and then it is divided into PS points and non-PS points again.PS points and non-PS points can be effectively classified through two parts of the operation without the need to set a threshold.By comparing the phase standard deviation of PS points of the improved method with that of the traditional method,it is confirmed that the PS points selected by the improved method have higher stability.