Dual Polarimetric HRRP Sequence Recognition Based on $H/\alpha$ Features

Dual polarimetric high resolution range prole (HRRP) is able to provide polarimetric and high resolution scattering characteristics information of radar target, which are very useful for automatic target recognition (ATR). In this paper, a novel feature extraction method, which combines both polarimetric and high resolution information, is proposed for dual polarimetric HRRP sequence based on H/α decomposition. In this method, the H/α decomposition is applied on HRRP sequence by temporal average instead of traditional spatial average, which retains the full resolution of HRRP data and provides efficient polarimetric characteristics information of target scattering centers. This paper also presents a scattering mechanisms classification method using the extracted H and α features, and analyzes the effect of noise and clutter on the two features as well as classification performance. The application and effectiveness of the proposed method are demonstrated via outfield experiments. Results show that scattering mechanisms and structural components of a volumetric target are recognized over high resolution range cells, and the proposed method is able to provide support for HRRP sequence-based radar ATR.