L. Du, H. Liu, and Z. Bao (PRC)
Radar Target Recognition, High-Resolution Range Profile(HRRP), Time-shift Invariant Feature, Higher-orderSpectra, Template Matching Method (TMM), RadialBasis Function Networks (RBFN)
Radar high-resolution range profile (HRRP) is very sensitive to time-shift and target aspect variation, therefore, HRRP based radar automatic target recognition (RATR) requires efficient time-shift invariant feature extraction approach and robust feature template establishment method with good generalization performance. A computational efficient method is proposed to calculate the Euclidean distance in the higher-order spectra feature space. This method is performed in HRRP space directly, which avoids calculating the higher-order spectra, so the storage requirement can extremely decrease. According to the widely used scattering center target model, the theoretical analysis and experimental results show that the average profile in a small target aspect sector has better generalization performance than the average feature vector in the same aspect sector. Finally, the recognition experiments of higher-order spectra features and original HRRPs are performed based on the measured data. The recognition algorithms include Template Matching Method (TMM) and Radial Basis Function Network (RBFN) classifier. The experimental results show that the power spectrum feature has the best recognition performance among the higher-order spectra features, including the well-researched bispectra feature.
Important Links:
Go Back
