Design And Implementation Of Multi-level User-Oriented Engineering Spaceborne SAR/InSAR Data Processing Software

With the characteristics of working in the circumstance of all day and all weather,Synthetic Aperture Radar(SAR)can acquire two dimensional microwave images of the land with high resolution.While the Interferometric SAR(InSAR)takes advantage of two SAR antennas to retrieve the Digital Elevation Model(DEM).Nowadays many SAR/InSAR systems have been applied into the civil or military applications.Due to the systems' large amount of data,complicated and highly professional data processing procedure,it is of great value to develop a SAR/InSAR data processing software,which is multi-level user-oriented,and has the ability of handling data in a large scale for either the whole procedures or a customized procedures.In this thesis,the architecture design and implementation of engineering SAR/InSAR data processing software are analysed in details.The main contributions are listed as follows:(1)The advantages and disadvantages of existing SAR/InSAR data processing softwares are summarised,and the necessity and feasibility of developing a multi-level user-oriented spaceborne engineering SAR/InSAR data processing software are expained.Based on the research into SAR/InSAR geometrics principle and data processing,the architecture of a multi-level user-oriented spaceborne engineering SAR/InSAR data processing software incorporating the engineering software development ideas are proposed.(2)Following the architecture design scheme,the user interface of the multi-level user-orientedand spaceborne SAR/InSAR engineering data processing software is designed based on engineering application requirements,and 5 core functions which are multi-level users management,high compatibility,large-scale batch processing,highly effective parallel-processing and high liability are developed.(3)To solve the problem of slow looking up and hard filtering of large-scale data,a hierarchical and classified displaying and management strategy based on tree structure is proposed.By exploiting modules such as large-scale data batch adding or deleting,data displaying or looking-up,data fusion or analysis,to achieve the software efficient management of large-scale data.(4)To solve the problems of low efficiency and bad precision of single-step processing of large amount of data,the following two strategies are proposed: large-scale batch processing of data and intelligently choosing of algorithm.The large-scale batch processing makes it possible to process multiple groups of data simultaneously using a single module,and the customized multiple procedures or the whole procedures can be processed alternatively in an automatical way,thus insuring high efficiency during processing.The intelligent choosing processing algorithm can make different algorithms and parameters available to different level users,and ensure high precision of the whole procedures with data characteristics and prior knowledge of the land cover.The performance of the proposed ideas which can satisfy the requirement of engineering application is validated by experiments based on Terra SAR-Xreal measurement data.