Influence of Non-parallel Trajectories on the Correlation of Unmanned Aerial Vehicle SAR Images

Due to the low cost and flexible operation of Unmanned Aerial Vehicle (UAV), UAV Synthetic Aperture Radar (SAR) is able to acquire the images of areas of interest in the case of disaster or emergency. Data from repeat tracks are very important because their capability of multi-dimensional mapping like interferometry or tomography. However, due to the characteristics of small size and light weight, UAV will inevitably be affected by airflow disturbance, resulting in non-ideal repeat tracks. Among them, non-parallel repeat trajectories are important factors to be considered while evaluating the correlation of SAR images. This paper studies the geometric decorrelation of SAR image pairs under non-parallel trajectories and deduces its mathematical model. The method is to use an equivalent synthetic aperture to approximately represent the non-parallel trajectories, and then select the non-orthogonal bases for decoupling according to the non-orthogonal direction of the envelope resolution. The derived mathematical model shows that the correlation of SAR image pairs is related to the azimuth angle of the equivalent synthetic aperture and the angle between the master and slave trajectories. Finally, the accuracy of the derived analytical model is verified by simulations and measured UAV SAR data.