GESTURE CONTROL OF A QUADROTOR UAV

Abstract

Natural interaction systems, particularly those based on computer vision, provide intuitive and effective communication between humans and drones. Gesture control allows for direct piloting of the drone, while person tracking ensures that the drone maintains visual contact with the operator. One function without the other is incomplete: gesture control is useless if the drone cannot see the user, and simple tracking offers no means of command. This project aims to combine these two functionalities by developing an intelligent system where a gesture recognition model and a person detection model work in tandem. The objective is to create a seamless, robust, and non-restrictive human-drone interface, where the user's body becomes the primary control device. Key words: Human-drone interaction,gesture recognition , person tracking, computer vision, drone control, natural interaction.