ABSTRACT

As a basic scattering element, arc-shaped structures can present considerable geometric information of the observed targets in synthetic aperture radar (SAR) images. The arc-shaped structures of various artificial objects sometimes completely disappear in radar images, which dramatically degrades the applicability of SAR images. Conventionally, multi-aspect observation is required to recover the disappeared structures, but this method is not always effective in some cases. This paper proposes a parametric reconstruction of arc-shaped structures in SAR imaging. The proposed method has the ability to reconstruct disappeared arc-shaped structures via three observations. The core idea is to use the anisotropic phase feature to distinguish the endpoints of arc-shaped structures from multiple point-like targets in the image domain and restore the arc through the scattering model. The method basically includes three procedures: arc-shaped structure recognition, parameter estimation, and target reconstruction. Computer simulations and a microwave anechoic chamber experiment are conducted to examine the effectiveness of the method.

摘要

作为基本的散射元素，弧形结构可以在合成孔径雷达（SAR）图像中呈现出被观察目标的大量几何信息。各种人造物体的弧形结构有时会在雷达图像中完全消失，大大降低了SAR图像的适用性。通常，需要多方面观察来恢复消失的结构，但这种方法在某些情况下并不总是有效。本文提出了SAR成像中弧形结构的参数化重建。所提出的方法能够通过三个观察重建消失的弧形结构。其核心思想是利用各向异性相位特征从图像域中的多个点状目标中区分出弧形结构的端点，并通过散射模型还原弧形。该方法主要包括弧形结构识别、参数估计和目标重建三个过程。进行计算机模拟和微波消声室实验以检验该方法的有效性。