Trajectory Follow And Obstacle Avoidance For Flight Safety On Micro Aerial Vehicle Using Monocular Camera

Abstract

This paper proposes an obstacle avoidance technique for micro aerial vehicles (MAVs) by using a monocular camera. The drone follows the trajectory for its mission as a sequence of waypoints. For avoiding an obstacle while the drone is processing the path, a sequence of images captured from the onboard camera that would be used to detect any object that might be facing. We use computer vision to determines the free zone along with the obstacle that may cause a crash and send feedback control to the drone. After that, the multi-rotor move to the safe area for avoiding an obstacle then resumes the first trajectory.