复杂目标雷达图像形成机理分析

合成孔径雷达(Synthetic Aperture Radar, SAR)图像包含了复杂目标重要的电磁散射特性，图像中的散射中心可以提供目标结构和散射点位置等重要的目标几何信息。定量分析SAR图像中散射中心分布和形成机理在目标探测和识别领域具有重要的理论意义和应用价值。本文选取典型复杂目标（航母和战车），正向构建了目标散射中心参数化模型。在此基础上，对目标SAR图像中轮廓的形成机理和图像稳定性开展了定量分析，建立了图像中散射中心与散射部件间的良好映射关系，揭示了散射中心的形成机理，进一步分析了目标上随姿态稳定的散射中心与目标局部几何结构之间的联系。此外，针对实际应用中目标几何模型非理想光滑的情形，本文研究了目标表面几何粗糙对SAR图像的影响，目标表面合适的粗糙程度可以改进仿真图像的结果，形成更加清晰的目标轮廓；但若粗糙程度过大则难以客观反映实际模型的电磁特性。本文构建的散射中心模型与实测数据进行了比对，结果吻合良好。

Analysis of Radar Image Formation Mechanism of Complex Target

SAR image contains the important electromagnetic scattering characteristic of complex targets, the scattering center in the image can provide important target geometric information such as target structure and scattering point position. Quantitative analysis of the distribution and formation mechanism of scattering centers in SAR images has important theoretical significance and application value in the field of target detection and recognition. This paper selects typical complex targets (aircraft carrier and combat vehicle) and constructs the parametric model of target scattering center in the forward methods. On this basis, the formation mechanism and image stability of the contour in the SAR images are quantitatively analyzed, the good mapping relationship between the scattering centers and the scattering components in the image is established, the formation mechanism of the scattering centers is revealed, and the relationship between the scattering centers stabilized with attitude on the target and the local geometry of the target is further analyzed. In addition, in view of the non-ideal smoothness of the target geometric model in practical application, this paper studies the influence of the geometric roughness of the target surface on the SAR image. The appropriate roughness of the target surface can improve the results of the simulation image and form a clearer target contour. However, if the roughness is too large, it is difficult to reflect objectively the electromagnetic characteristics of the actual model. The scattering center model constructed in this paper is square with the measured data.