ISAR Image Reconstruction of Accelerated Motion Using Matching Pursuits

The problem of inverse synthetic aperture radar (ISAR) image reconstruction of accelerated motion is considered. As a result of accelerated motion experienced by the target during the image integration period, the ISAR image is distorted and the Doppler frequency of the radar echoes produces undesired change. In traditional ISAR imaging, uniform motion is usually assumed. When accelerated motions are present, it is not possible to obtain a focused image of the target. To solve this problem, the basic idea is to exploit the characteristics of the Doppler frequency of every range cell. The radar signal is decomposed by matching pursuit with a three-parameter chirplet atom. Different point scatterers in cross direction can be separated in the parameter space. The proposed approach is applied to the point scatterer simulation and the result demonstrates that it can improve the smeary ISAR image better than what can be achieved by conventional Fourier transform methods and Winger-Ville Distribution.