Research On Multi-baseline Spaceborne InSAR Imaging And Data Pcocessing

Multi-baseline spaceborne InSAR technique uses multiple interferometric phase maps corresponding to different baselines of the satellites to perform elevation inversion.Its advantages are that it can effectively reduce the random error caused by single baseline interferometry,and can overcome the adverse effects caused by phase stacking,abrupt terrain,noise and other factors.The technique can greatly improve the accuracy of DEM for complex terrain,which is of great significance of research and application.This thesis conducts research on multi-baseline spaceborne InSAR imaging and data processing.The main work and innovation of the thesis are as follows:1.Constructs the multi-baseline spaceborne InSAR imaging model with one transmitter and four receivers.Based on the knowledge of satellite orbit parameters,satellite-earth coordinate system transformation and the InSAR theory,the multi-baseline spaceborne InSAR imaging simulation model is constructed,which takes the orbit conversion,earth rotation effect,satellite-to-earth geometry and baseline coupling into consideration.The model provides the basis for subsequent simulation experiments.2.A cascade iterative multi-baseline spaceborne InSAR surface BP imaging algorithm based on the long and short baselines is studied.The BP algorithm owns the high applicability to arc trajectory of satellite due to its accurate distance calculation,and can achieve coarse registration of the SAR images and flat-earth phase removal.Aiming at the compressed terrain of the observation scene during the imaging process and the interference phase loss due to the phase stacking effect,the surface BP algorithm is introduced and combined with the multi-baseline InSAR technology and the advantages of long and short baselines are well adopted.Due to the high ambiguity of the short baseline,the interference phase fringes are sparse and easy to unwrap.The elevation derived from the short baseline interferometry is used as the reference terrain information to construct a new surface imaging space.Then the echo data corresponding to long baselines are projected to this new surface space.This method can avoid the problem of dense phase fringes,reduce the difficulty of phase unwrapping and achieve higher accuracy of DEM.The effectiveness of the method is verified by simulation tests.3.A multi-baseline spaceborne InSAR DEM fusion algorithm based on guided filtering is proposed.Based on the probability distribution characteristics of the interference phase,a method of calculating DEM using the height error to weighted average the height information of each baseline is studied.This method has insufficient ability to maintain terrain details and is susceptible to impulsive noise.For this problem,a multi-baseline InSAR DEM fusion algorithm based on guided filtering is proposed.The DEMs derived from each single baseline are decomposed into the base part and the detail part.The prior terrain is utilized as the base part to imply the general topographic trend.The detail part extracted by DEM decomposition is processed by guided filter with amplitude image of the original InSAR scene,which can well preserve the details of the actual elevation.Then the detail parts of each baselines get weighted averaged by the height error corresponding to different baselines.Finally,the addition of the base part and the detail part after the guide filter is the fused DEM.This method can fully combine the prior terrain information and the amplitude information of the original scene,which can well preserve the details of terrain and improve the DEM accuracy.The effectiveness of the proposed method is demonstrated by the repeat-pass multi-baseline TerraSAR data.The recognized high-precision LiDAR DEM ia taken as the reference of the actual terrain.Compared with the traditional weighted average method,the details of valleys and ridges in study area are preserved effectively and the accuracy of DEM is higher.