Image-based visual servoing of rotorcrafts to planar visual targets of arbitrary orientation

Abstract

This paper for the first time extends the virtual camera image-based visual servoing (IBVS) scheme to enable an underactuated rotorcraft UAV to regulate its translational motion and heading relative to a planar visual target of arbitrary orientation. The conversion from real camera images to virtual camera images of visual targets are proposed based on a set of rotation matrices. Hence, image moment features can be reused due to the simplicity and decoupled structure of the interaction matrix, and satisfactory 3D Cartesian trajectory of UAVs. In the design of the IBVS control law, the external disturbance and model uncertainties are estimated by an integral-based filter. In addition, to enable tracking of a moving visual target, a velocity estimator is developed. The global asymptotic stability of the error dynamics is proven. Both of the simulation and experimental results of tracking of a tilted moving planar target have been provided to show the efficacy of the proposed IBVS scheme.