In this paper a vision-based tracking controller is designed for the quadrotor vertical take-off and landing of an unmanned aerial vehicle. An imaged-based visual servoing approach is utilised to localise the quadrotor with respect to a moving target. Perspective image moments are used to define the visual features, which are projected on a rotated image plane to simplify the image dynamics. Attitude information and angular velocities are assumed to be available and the controller uses the flow of image features as the linear velocity cue. Presence of delay in processing and communication is modelled as a constant time delay in the force input of the translational dynamics, where a controller is designed for theses dynamics to compensate the delay effect. This controller is saturated in order to meet the quadrotor model constraint. A dynamic surface control approach is utilised for the rotational dynamics to track the desired attitude, defined through the position control loop. The stability properties of the complete control scheme are analysed using a theory of nonlinear cascaded systems. Simulation examples are provided in both nominal and perturbed conditions which show the effectiveness of the proposed theoretical results.