AN EFFICIENT ISAR IMAGING METHOD FOR MULTIPLE TARGETS

Abstract

This paper proposes an e-cient method to obtain ISAR images of multiple targets ∞ying in formation. The proposed method improves the coarse alignment and segmentation of the existing method. The improved coarse alignment method models the ∞ight trajectory using a combination of a polynomial and Gaussian basis functions, and the optimum parameter of the trajectory is found using particle swarm optimization. In the improved segmentation, the binary image of the bulk ISAR image that contains all targets is constructed using a two-dimensional constant false alarm detector, then the image closing method is applied to the binary image. Finally, the connected set of binary pixels is used to segment each target from the bulk image. Simulations using three targets composed of point scattering centers and the measured data of the Boeing747 aircraft prove the efiectiveness of the proposed method to segment three targets ∞ying in formation.