This article presents a comparison of three control techniques: nested saturations, backstepping, and sliding modes. The control objective consists of obtaining the best control strategy to stabilize the position of a quad-rotor unmanned aerial vehicle (UAV) when using visual feedback. We propose a vision-based method to measure translational speed as well as the UAV 3D position in a local frame. The three selected controllers were implemented and tested in real-time experiments. The obtained results demonstrate the performance of such methodologies applied to the quad-rotor system.