Obstacle Detection and Avoidance based on 3D Stereoscopic Technique for Autonomous Drone

Autonomous drones are drones that can fly on their own without the need for human supervision. Once the drone is assigned a flight mission, it will automatically fly to its destination. The ability to quickly maneuver obstacles is essential to complete the task assigned to it. Obstacle analysis processes need to be efficient and require fewer computing resources due to the hardware and power constraints of drones. This work presents a simple but effective way to detect obstacles using a 3D Stereoscopic camera and a less complicated image processing method, which can run on a small single-board computer. On-site flight tests show that the proposed method can be used efficiently. It can detect obstacles at a distance of 2 meters, 5 frames per second, 79.55% of detection accuracy, and maneuver effectively.