Brain-computer interface for AR at a low cost

This work presents the design, implementation and testing of a Brain Machine Interface (BCI) system based on µ waves to control the navigation of a commercial drone. BCI systems translate brain signals into commands that can be used to activate and control external applications. The µ rhythm is a type of brain response that is modulated through motor activity and can be easily measured using electroencephalography (EEG). For this reason, BCI systems based on µ waves have been extensively explored in the literature as a way to allow patients with compromised neuromotor systems to interact with the external environment. In this work, a software platform was developed to implement signal processing and application interface routines. Well-established techniques such as the CSP spatial filter and the LDA classifier were used to detect brain patterns. Furthermore, a methodology is proposed to translate the classifier output signal into commands that can be directly sent to the drone. To acquire the EEG signals, a low-cost and open-source amplifier called Open-BCI was used. The system implementation was validated using a public data set, which was used on the platform as a way of simulating the real-time behavior of the system. The application tests were conducted in a drone simulator, which demonstrated the correct functioning of the proposed methodology and the developed system.