Research On Multi-baseline SAR Interferogram Filtering

InSAR technology,a kind of earth observation technology,has played an important role in the investigation and monitoring of various geological disasters.It has been used in deformation monitoring,land classification,glacial movement,fault inversion,and landslide monitoring.However,because InSAR is a coherent measurement method,which needs multi-period SAR data to produce an interferogram.The influence of the factors such as decoherence are the main problem of InSAR deformation monitoring.The quality of the SAR interferogram determines the reliability and availability of the subsequent processing.The deformation information of the original interferogram is often submerged in the noise.Therefore,the interferogram denoising plays an important role in InSAR technology.It directly affects the phase unwrapping,DEM and deformation results.Therefore,this paper mainly takes the interferogram filtering as the research content,obtains the target of the bottom coherent region monitoring point,and finally carry out a series of research work based on the experimental data of farmland area.The main contents of this paper are as follows:(1)The development and main methods of SAR interferogram filtering are introduced in detail,and their own characteristics are divided into three levels: filtering in spatial domain,filtering in transform domain,and filtering in spatial-temporal domain.(2)I have studied the widely used Goldstein filtering algorithm.Through various filter parameters experiments,fianlly it is proved that the performance of this method is not good in the low coherent region.At the same time,the unbiased estimation of the coherence of the interferogram is explained and experimentation.(3)The advanced multi baseline SAR interferogram filtering method are introduced in detail,such as the eigendecomposition based on covariance matrix and the eigendecomposition based on the coherent matrix being carried out.(4)An improved SAR interferogram denoising method based on principal component analysis and the Goldstein filte is proposed.This method,especially in the low coherent region,can increase the density of the point and enhance the coherence.It provides a reliable filtering method for the difficult monitoring area.(5)In this paper,an interferogram based on coherent matrix eigendecomposition batch filtering algorithm is proposed.And the same time a kind of interference pair selection strategy is put forward,which improves the situation that the traditional small baseline(SBAS)is easy to leak and misselect interferences.Finally,this method is applied to the ground subsidence monitoring in Qingxu,where most of the ground surfaces are covered by farmland.The results show that this method is more effective than the improved Goldstein filtering,and the effective monitoring point target is obviously increased.