In this thesis the potential of visual servoing techniques for vertical take-off and landing(VTOL) UAV platforms is explored. An overview of UAV applications and state-of-the-art quadrotors is provided, backed up by the necessary theoretical background. Visualservoing based solutions of the important flight tasks (such as vertical take-off and land-ing, obstacle avoidance and attitude estimation) are presented.A novel vision based autopilot for vertical take-off and landing is developed, exper-imentally verified and implemented on a medium-size outdoor quadrotor platform pro-vided by Dr. YangQuan Chen from Utah State University, Center for Self-OrganizingIntelligent Systems.