Gaussian process-based visual pursuit control with unknown target motion learning in three dimensions

In this paper, we propose an observer-based visual pursuit control integrating three-dimensional target motion learning by Gaussian Process Regression (GPR). We consider a situation where sensor equipped rigid body pursuits whose velocity is unknown but dependent on the target's pose. estimate pose from information and (GP) model to predict estimate. analyse stability of proposed showing that estimation errors are ultimately bounded with high probability. Finally, simulations illustrate performance schemes even if measurement corrupted noise.