Research On Optimization And Improvement Of DORIS Software

InSAR/DInSAR processing is complicate and requires professional software, e.g.,GAMMA, Earthview-InSAR, and DORIS. DORIS is one or the most popular and opensource software, which is written in C++by Kampes et al. in the Delft University ofTechnology of Netherlands, aiming for studying the topography and surfacedeformation. Because of DORIS is free, the researchers can use or modify the codesaccording to their needs. However, the researchers may encounter some difficultywhile using DORIS. The software mainly runs on the UNIX/Linux operating systemconducted with the full command line manner. However, when Windows users need touse it, they have to run it under the Cygwin environment with the same manner asLinux users do. Utilizing DORIS in the Linux environment, the user does not has thegraphic interface, and also needs some specific tools (e.g., cpxfile) to check theintermediate results (e.g., SAR images or graphs). Since the common users areaccustomed to the Windows and are not able to use Linux operating system for the lackof the basic knowledge of the Linux system. Also, they have no chance to check theintermediate results directly. In addition, some algorithms designed in DORIS are notso perfect, e.g., its baseline estimation algorithms need to be optimized and modified toimprove the precision and accuracy.Therefore, it is helpful to design and develop thecrossing-platform DORIS software with friendly interface and implement some noveland optimized algorithms to replace the olds.To this end, based on a systematicexamination of the DORIS software structure and relative knowledge, this researchmainly concentrated on the followings:First, designing and implementing a newly and optimized baseline estimationalgorithm to replace the old. In the new algorithm, it mainly considers the fact that theearth reference ellipsoid radius at different latitudes is difference. In theimplementation of the new algorithm, the calculation of ellipsoid radius is morerealistic, which can remarkable reduce the estimation error, and improve the accuracyof the baseline estimation. Compared the extracted DEM results through the2baselineestimation algorithm with the ASTER GDEM (referred object) respectively, it shows that the absolute error of the improved baseline estimation algorithm decreased0.1m,and its relative error reduced1.1%.Second, a specific interface system, WinDORIS, has been designed and developed.WinDORIS has the capability to overcome the crossing-platform problems withClient/Server (C/S) architecture, and has a friendly interface with wxPython. UnderWinDORIS, the Windows users can easily to use and also has a perfect performancecompared with it runs in the Linux environment.Meanwhile, pointing to the problem that DORIS cannot support the processing forALOS/PALSAR data at present; we use the approach of modifying its intermediatefiles to make the verification experiments process smoothly.Finally, we finished the verification experiments by DEM extraction, two-pass andthree-pass method using the data of multiple sources, multiple SAR product formatsand multiple experimental areas. Compared time cost with DORIS, and the resultshowed the time cost in WinDORIS was longer than that in DORIS by0.1s. From theverify experiments, it can be proved that it is satisfactory to processing SAR data usingWinDORIS software, which can not only provide a convenient, fast-response anduser-friendly GUIs; but also obtain a correct and effective results.At present, WinDORIS has been tested and obtained satisfactory results, whichwill be helpful and valuable for Windows users to use the open source software andalso play a significant role to software designing and development utilizing C/Sarchitecture.