Development of a UAV with hand gesture recognition using deep learning

• Main Author: Ng, Calvin Alexander Y.
• Format: text
• Language: English
• Published: Animo Repository 2019
• Subjects: Drone aircraft—Control systems; Gesture recognition (Computer science); Mechanical Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/6393; https://animorepository.dlsu.edu.ph/cgi/viewcontent.cgi?article=13428&context=etd_masteral
• Institution: De La Salle University

Existing Unmanned Aerial Vehicles are typically controlled by the use of Radio Control which is a specialized device that translates button presses and joystick commands into movement. This means that control of an Unmanned Aerial Vehicle is both susceptible to radio interference, and is unintuitive. Recent research has proven that hand gestures are the most intuitive method for quadcopter control. However, past research has always used ground-based computers to perform gesture recognition algorithms, which means that the overall system is still susceptible to electromagnetic interference. This thesis presents the development of a quadrotor Unmanned Aerial Vehicle that uses an onboard companion computer to achieve gesture recognition with a deep learning algorithm without the need for a ground-based computer.