Hand gestures recognition using machine learning for control of multiple quadrotors
David Valencia Redrovan, Donghan Kim · 2018
Interacting with a gesture is natural and simpler than manipulating physical devices or controls, based on this, in this paper we propose the design of real-time hand gesture recognition for flight control of multiple quadrotors through electromyography signals (EMG) and Convolutional Neural Network (CNN) in order to simplify flight operation control and make it more intuitive for the user. Additionally, a Sliding Mode Control(SMC) algorithm is implemented to control the quadrotors formation during flying, based on leader-follower principle. The results demonstrate the effectiveness of the proposed interactive method.