This paper focuses on the domestic probe-drogue aerial refueling,and designs a set of vision-based autonomous aerial refueling system for drogue tracking and locating.Firstly,the tanker system is modeled.The finite element analysis is used to model the refueling hose drogue,which provides reference data for the movement of the drogue.A two-dimensional image localization of the drogue is completed by using the minimum ellipse localization,and the LHM algorithm is used to complete the three-dimensional localization.Then,considering the poor stability and low output frequency,EKF is introduced to improve the robustness of the system and decrease the time for image processing.Finally,the simulation results show that the vision-based measurement using EKF can improve robustness and output frequency under different turbulence conditions.