A Vision-Based Guiding Method for Autonomous Route Planning of Micro Quadrotor UAV

In this paper, a vision-based guiding method was proposed to address the issue of autonomous route planning for micro Quadrotor Unmanned Aerial Vehicle (quad-UAV). In the autonomous flight stage, the skeleton line, which represents the main extension of roads, is extracted to get the potential forward direction (FD). Then the current FD is defined as the minimum value of all angles between potential FD with the last FD. In the stage of autonomous landing, the parking apron is detected based on the roundness of a closed region with an appropriate threshold. To tradeoff the stability and speed, the forward destination location is calculated according to the amplitude of change of the FD. Additionally, the PID controller is further applied in the whole route planning. The proposed algorithm is verified on the parrot micro quad-UAV platform. And the results illustrate that the proposed scheme can autonomously control the UAV fly along the road and land on the parking apron, and exhibit good performance.