Abstract

Target identification using Inverse Synthetic Aperture Radar (ISAR) imaging is an important tool required with the current day high-resolution radars for threat assessment and evaluation. Identification of target using the ISAR images is gaining popularity because of its importance in decision-making. The ISAR image represents the target’s reflectivity measured at discrete frequencies in the RF domain. It uses the Doppler information to obtain the cross range resolution, and the range resolution is directly related to the bandwidth of the transmitted signal. In order to capture the target information effectively, the Wigner-Ville Distribution (WVD) based ISAR imaging was proposed. But this involves high computational complexity and hence difficult to realize for practical applications. In order to reduce these computations and make it feasible for practical scenarios, the Harmonic Wavelet (HW) based ISAR imaging is proposed and the resulting images are used for target identification. The performance of the proposed scheme is compared with the Filtered Wigner-Ville Distribution (FWVD), standard STFT and FFT techniques, and found to provide better ISAR images with 92% reduction in computations. A Neural Network based Automatic Target Identification (ATI) scheme invariant to translation, rotation and scale is used for target classification.