Avian Dynamic Electromagnetic Scattering Feature Extraction Based on Wavelet Transform and Singular Value Decomposition
Radar bird detection is a hot issue in the fields of aviation safety and environmental ecology.The radar scattering cross section(RCS)of bird targets is small and the scattering features are single,which brings many challenges to bird detection.To solve these problems,a dynamic electromagnetic scattering feature extraction method based on wavelet transform and singular value decomposition for flying bird targets is proposed.Firstly,the dynamic RCS sequence of the flying bird with flapping frequency of 2-20 Hz is wavelet transformed to obtain the wavelet coefficients of each branch,and then the wavelet coefficients of each branch are reconstructed.The feature matrix composed of wavelet coefficients is decomposed by singular value to quantitatively describe the dynamic electromagnetic scattering features of the flying bird with eigenvalues.To verify the validity of the method,numerical experiments with incident frequencies of 0.5,1,and 3 GHz are conducted in this paper under circling and leveling trajectories,horizontal polarization,and vertical polarization,respectively.The results show that the eigenvalues show obvious linear correlation with the flapping frequency of the flying bird,which can effectively reflect the motion characteristics of the flying bird and provide new perspectives and ideas for the radar detection and identification of bird targets.
wavelet transformsingular value decompositiondynamic radar cross sectionflapping frequencyfeature extraction
国家科技期刊平台
NSTL
万方数据
