Typical Target Discrimination Using New High-Resolution Polarimetric SAR Data

In 2007,three new high-resolution SAR satellites COSMO-SkyMed, TerraSAR-X,RADARSAT-2 have been successfully launched,which provide new data for the remote sensing fields worldly and have aroused close interest of scholars and experts both at home and abroad.It is hoped that the new data sources can inject new blood into applications in various fields of the national economy.Of course,the improvement of resolution and the enrichment of polarization have provided possibility for the successful application of SAR data in various fields.This paper devoted on the more effective recognition effect of typical features using the new high-resolution polarimetric SAR data.several aspect analysis have done based on the new sources:1.In view of new high resolution SAR data characteristic,the author did analysis of high resolution SAR image characteristics of typical features and analyzed the typical features in accordance with the provisions of the second survey of the first broad categories:land,construction land,forest,grass,transport land,water or water conservancy facilities and also did detailed analysis of the structural characteristics and the property of scattering of the typical features,as well as the imaging characteristics in different polarization.2.Introduced high resolution SAR image pretreatment method,radiometric calibration,geometry correction as well as the SAR image filtering is the most main three aspects.Considering the shortage of optical data in cloudy rain area,Proposed the optimized choice basis When geometry correction between SARs.And has compared the commonly used filter method regarding the high resolution SAR data filter effect.3.The different mode of high resolution SAR data has different structural feature as well as detail attribute ability.This paper did comparison about typical features recognition effect of different mode of SAR data and also made comparison about data enhancement and features recognition effect of high-resolution multi-temporal SAR data.Then carried on the multi-temporal SAR data combination classified effect appraisal take Beijing Tongzhou District as the test area.4.Polarization information enhancement greatly improved the ability to identify the typical features.In this paper,we selected Zhazuo forest farm as the test area and made comparison of backscattering value of typical features on full polarization RADATER-2 images.The author used methods of polarimetric target decomposition, commonly used at home and abroad,to analyze the identification of typical features. Then make comparison of polarization response flat of typical features as well as different parameters of polarization decomposition.5.In view of commonly polarimetric classification method based on polarization diffusion properties and target decomposition theory,this paper carry on the classified effect comparison base on the Zhazuo forest farm as the test area.In order to provide the suggestion of typical target discrimination using polarimetric SARs.