Wideband Radar Target Polarimetric Features Extraction And Recognition

Radar target recognition is an important problem needed to be solved urgently for modern radar to implement target position, measurement and recognition. However, because of the complexity of the problem itself, there is still no satisfying solution. The combination of high resolution and full polarization is considered as the one of the most promising approach to solve the problem, and thus high-resolution fully polarimetric radar is also becoming an important direction of modern radar development. In this dissertation, wideband radar target polarimetric features extraction and recognition for the high-resolution fully polarimetric radar system is discussed in depth, mainly including fully polarimetric scattering center parameter estimation, wideband radar target polarimetric feature extraction and recognition based on the polarimetric target decomposition and distribution of scattering centers. The main work in this dissertation is organized as follows:The first chapter introduces the significance and status of wideband radar target polarimetric feature extraction and recognition, and gives the work profile of this paper.The second chapter describes the characterization polarimetric scattering characteristics of radar target , and a method combinating polarimetric target decomposition and scattering centers estimation is proposed to analyze polarimetric scattering characteristics of radar target. This method is validated by real data measured in anechoic chamber.The third chapter firstly introduces the characteristics of fully high range resolution profile(HRRP) and polarimetric target decomposition theory. Subsequently, it applies the Krogager decomposition, Cameron decomposition and Cloude decomposition to the fully HRRP of radar target to extract three types of polarimetric features, based on which a target recognition procedure is designed; Finally, two simulation experiments on data of two groups of ships calculated through the electromagnetic software are carried out to test the effectiveness of these polarimetric features. Compared to the recognition performance of the feature extracted from single polarization HRRP, the method in this paper effectively improves the recognition performance by employing fully polarimetric information.The fourth chapter firstly works on fully polarimetric scattering center parameter estimation using Unitary ESPRIT according to the location consistency of scattering centers in different polarization channel. Unlike the conventional scattering center estimation under single polarization channel, the proposed method can not only estimate all parameters in one operation with high estimation accuracy, but also avoid such problems as screening and matching of scattering centers, which greatly reduces the computation burden and improves the processing efficiency. Secondly, for ship targets, given the sparseness differences of scattering centers distribution, a target recognition method based on sparse degree of scattering centers is put forward. In fact, it is observed from simulation experiment results that the feature of sparse degree can provide good recognition performance, which proves that the method is reasonable and effective.The last chapter presents a conclusion of this dissertation and a description of the future work.