Accurate 3-dimensional image reconstruction algorithm extending RPM method to ISAR model

摘要

Microwave 3-dimensional reconstruction techniques exploiting multiple ISAR (Inverse Synthetic Aperture Radar) images have been well established, and are suitable for the recognition of artificial targets such as aircrafts or ships. However, the existing algorithms assume only an aggregation of multiple points, and most require the tracking of the multiple points over multiple ISAR images. In the case of a solid object with a continuous boundary, such as a wire or polyhedral structure, these algorithms severely suffer from the inaccuracy for 3-D imaging, owing to scattering centers shifting on the target surface in terms of the observation angle. To tackle the above difficulty, this paper extends RPM (Range Points Migration) method to the ISAR observation model, where a double mono-static model is introduced to suppress cross-range ambiguity. Numerical simulation demonstrates that the proposed algorithm has a distinct advantage in accurate 3-D imaging, even for not point-wise targets.