Research Of An Improved Phase Unwrapping Algorithm And Integration Development Of Doris Software

Synthetic aperture radar interferometry(Interferometric InSAR, the radar interferencemeasurement) is one kind of technology to get digital elevation model (DEM) by phaseunwrapping after acquiring interference images by calculating the phasedifferent of twoimages that based on the same region. Interferometric radar has become a hot topic in thefield of geologists due to its all-weather, all-day and penetrating features.Delft Object-oriented Radar Interferometric Software(Doris) is a set of open source,cross-platform InSAR data processing systemwhich can process ENVISAT, JERS, ALOSand other satellite sensor data. The Doris has very powerful radar data processingcapabilities, but the human-computer interaction interface is ugly and difficult and theparameter configuration process is cumbersome and need using command line toimplement one-step data processing and need text editing the parameter file, these defectslead to check the intermediate or final processing results very inconvenience.It is very meaningful to design and develop a VDoris(Visual Doris) in order toovercome the defects of Doris software. The main system research purpose is to provideusers with SAR data processing function module selection, process validation, processexecution, parameter configuration module, image display and friendly interactive interfacebased on Doris system powerful core processing functions and provide a friendly visualdata processing system to non-professional remote sensing field experts by implementingthe transfer and execution through business logic management. This paper has researchedthe system structure and module functions of Doris in the light of the purpose of this articleand analyzed the advantages and defects of Doris and researched the data processing ofDoris software modules and related configuration files for the effective segmentation. Thispaper has proposed a project that can create modules such as the base class for all childfunction and make the system having better stability and extensibility, eventually make theprocess function module selection and parameter configuration simple and functional easyto view performance data processing and the result images visual by analyzing the commonoperations and properties of all sub-functions.Wrapped phase solution treatment is the key of the InSAR data processing, whetherthe bound algorithm can meet the requirement for precision or not directly affect thegeographical elevation model established.Chenproposed network flow algorithm in2001, the algorithm gaveattention to both the operation efficiency and accuracy of the results.However, there are still two problems by the use of this algorithm on rugged areas afterfiltering the interference figure:(1) can’t get results all the time;(2) the solution results ishighly affected by the data coherence factor. It is worse in vegetation cover, the air humidareas, coherent factors. This undoubtedly hindered the application of InSAR technology inrugged terrain area.This paper has proposed an improved algorithm based on Snaphualgorithm using the mask method in order to overcome the defects that is using coherentdata probability distribution method to control the low coherence pixels not to attend thesolution of the sling operation.The above step can control the error propagation andapplying digital elevation surface fitting method to compensate for low elevation points.Last the experimental demonstrated that the improved algorithm having a perfect effect.