Vision-based control of a quad-rotor UAV

Abstract

This chapter introduces two different vision-based control strategies for stabilizing a quad-rotor during flight. The first strategy is based on a homography estimation technique and an optical flow computation. Using this approach, a comparison of three control methods is addressed, with the purpose of validating the most effective approach for stabilizing the vehicle when using visual feedback. In the second strategy, the vision system is implemented for allowing altitude control, which allows stabilizing the 3-dimensional position and regulating the velocity of the vehicle using optical flow. For validating the effectiveness of the two vision-based control strategies, the tasks of autonomous hover and navigation are executed in real-time experiments.