EMG Based Control of a Quadcopter

In this paper, we present a work on control of quadcopter using electromyography (EMG) signals acquired from subject's muscles. A simple yet effective interface is maneuvered that can permit even amateur users to control a quadcopter. EMG signals were acquired using eight electrodes of MyoArm band while the subjects performed four different hand gestures to control the quadcopter in four different directions. Moreover, those signals were filtered to remove noises-the noise-free signal was then used to acquire features that could be used for classification of different hand gestures. After acquiring suitable features, linear discriminant analysis was used to classify and generate control commands to control the quadcopter. Results demonstrate the feasibility in using EMG to control a quadcopter effectively with little training.