Micro‐Doppler effect removal in inverse synthetic aperture radar imaging based on UNet

Abstract

AbstractThe micro‐Doppler (m‐D) effect caused by the rotational parts of the targets influences the quality of inverse synthetic aperture radar (ISAR) imaging. In this letter, a novel deep network‐assisted method is proposed to reduce the m‐D effect in ISAR imaging. The training data, including ISAR images with m‐D effect and ISAR images without m‐D effect, help the network establish non‐linear mapping relationships. The simulated and measured data results show the effectiveness of the proposed method.