Designing a Drone Controller using Electromyography Signals
Yash Doshi, Divyanshi Nath · 2021
The aim of this project is to introduce new-fangled way to control Unmanned Aerial Vehicle (UAV) using electromyography (EMG) signal. Electromyography signals are biopotential signals that are generated when any muscular action takes place in a person's body. The idea is to provide a novel method of controlling drones, and other remote-controlled objects by recognizing hand gestures using EMG Signals. This paper presents a controller designed that can be easily fitted to a band on the user's arm. It is composed by inertial (gyroscope, accelerometer) and electromyographic sensors, that will measure the muscle movements and construe them in references for the drone. Gesture recognition is achieved using an Artificial Neural Network to fully exploit the capabilities of both, the user and UAV - Unmanned Aerial Vehicle. UAV chosen to test with the controller is a quadcopter. The comparative results show the compactness and easy-to-handle interface that allows even inexperienced users, to intuitively have control over remote-controlled objects as compared to a standard remote controller.