Font Size: a A A

Research On Polarization Radar Sparse Imaging And Feature Extraction Method

Posted on:2017-02-04Degree:MasterType:Thesis
Country:ChinaCandidate:D SunFull Text:PDF
GTID:2428330569998966Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
High resolution is an important direction for the development of Synthetic Aperture Radar?SAR?.The conventional SAR imaging algorithms based on matched filtering is limited by the signal bandwidth and the synthetic aperture length.One of the important aims of SAR image processing is to realize the classification and recognition of targets.A feasible method is based on acquiring the target's feature parameters.In recent years,sparse signal processing theory-Compressive Sensing?CS?theory has attracted the attention of scholars both domestic and overseas.Electromagnetic scattering signals of man-made objects such as vehicles and buildings have obvious sparse characteristics,and the total electromagnetic scattering of the target can be approximated by the synthesis of local scattering's center scattering.Therefore,CS has great potential in improving the quality of radar imaging and for high precision imaging of man-made targets.In addition,the introduction of CS to the inverse scattering problem is physically feasible.However,the application of CS theory in radar imaging and feature extraction is a new field,it is still in the initial development stage with many problems need further study.Based on the polarimetric SAR imaging radar system,this paper deeply studies the sparse target polarization high resolution imaging method,then inverses the fine geometric structure and size of the sparse target.In this paper,we propose a sparse image reconstruction method based on sparse reconstruction,an adaptive model parameters selection method and polarization geometry feature extraction method.The main contributions of this paper are listed in the following three aspects:1.The sparse reconstruction theory is summed up,the 1l-norm minimization,greedy,lp-norm minimization and Bayesian methods are studied,and a typical algorithm in sparse reconstruction is selected as the background to simulate the experiment.The performance of each method is compared and analyzed from reconstruction precision,super-resolution and SNR sensitivity,respectively.2.The imaging methods of sparse target radar are summarized.For the problem of the huge computation amount or on certain model relying of parameter selection methods,the relationship between the model parameters and the signal variance and the noise variance is deduced,and an adaptive parameter selection method based on the maximum posteriori probability estimation and the Bayesian reasoning is proposed.Finally,based on one-dimensional simulation data and two-dimensional measured data,the validity of the proposed method is verified.3.The feature extraction method of target polarization geometric structure based on sparse reconstruction is studied.Firstly,the attribute scattering center model and the scattering center model are introduced.A parameter estimation method based on sparse reconstruction and its concrete realization steps is given.On the basis of the simulation data,the lp-norm method which can adaptively determine the model parameters and other two sparse reconstruction methods are used to estimate the parameters of the scattering center respectively.The experimental results show that the lp-norm method has the most accurate performance.
Keywords/Search Tags:Synthetic Aperture Radar(SAR), polarization, compressive sensing, sparse reconstruction, parameter selection, feature extraction
PDF Full Text Request
Related items