Research On Deformation Extraction Algorithm Oftransmission Tower Based On SAR Tomography

SAR tomography(TomoSAR)is an emerging three-dimensional(azimuth-distance-elevation)imaging technology that overcame the problems of top-bottom inversion,shadow and layover in traditional SAR imaging technology,and met the strong demand for 3D information in application areas such as surface detection and city target recognition.It became an important developmet direction in the field of microwave remote sensing imaging.Differential SAR tomography(D-TomoSAR)is an extension of SAR tomography,synthesizing the aperture at the deformation rate direction(line of sight)to obtain the resolution of the deformation rate.Therefore,D-TomoSAR achieves four-dimensional imaging.In this paer,the transmission tower is taken as the research target.The elevation and deformation rate information and the inclination displacement extraction of the top of the transmission tower are studied.The main work of this paper is as follows:1)Studying the imaging principle of SAR tomography and Differential SAR tomography.According to the structural characteristics of the transmission tower,this paper studies the PS point identification algorithm to screen out the pixels of subsequent imaging.The performance of the amplitude deviation method and amplitude threshold method are analyzed by the real SAR data.The results show that the amplitude threshold method is better than the amplitude dispersion method.Meanwhile,in order to identify the scattering point information in the same pixel,three algorithms AIC、MDL、GDE are simulated respectively.AIC and MDL algorithms perform better under low SNR,GDE performs poorly.2)Studying the three D-TomoSAR algorithms based on Capon algorithm,OMP algorithm and SWOMP algorithm.According to the the large side lobes and false imaging results in the SWOMP algorithm,a SWOMP algorithm based on secondary screening is adopted.This algorithm removes the interference of the false target by the priori information of the number of scatterers point and secondary screening,the side lobes is smaller.Then,According to the simulation experiment,the performance analysis of the above four imaging algorithms are carried out.The information of the transmission tower was extracted by TerraSAR-X data from Maoxian County and Xichang area of Sichuan Province,and elevation and deformation rate information of the reconstructed position were obtained.It confirms the feasibility of Differential SAR tomography on transmission towers.Finally,using real SAR data to extract information of bnuildings.3)Studying the imaging principle of DInSAR technology.Then,studying DInSAR model by comparing the inclination displacement of the transmission tower with the surface deformation.According to the geometric relationship between the radar satellite and the transmission tower to derive the formula for calculating the differential interference phase caused by the inclination displacement of the top of the transmission tower.Thus,using the imaging principle of DInSAR,the inclination displacement of the top of the transmission tower can be calculated by obtaining the differential interference phase at the top of the transmission tower.Finally,comparing SAR data with field measured data,the validity and applicability of the model are verified.