A fully-wearable non-invasive SSVEP-based BCI system enabled by AR techniques for daily use in real environment.

Abstract

This thesis aims to explore the design and implementation of Brain Computer Interfaces (BCIs) specifically for non medical scenarios, and therefore to propose a solution that overcomes typical drawbacks of existing systems such as long and uncomfortable setup time, scarce or nonexistent mobility, and poor real-time performance. The research starts from the design and implementation of a plug-and-play wearable low-power BCI that is capable of decoding up to eight commands displayed on a LCD screen, with about 2 seconds of latency. The thesis also addresses the issues emerging from the usage of the BCI during a walk in a real environment while tracking the subject via indoor positioning system. Furthermore, the BCI is then enhanced with a smart glasses device that projects the BCI visual interface with augmented reality (AR) techniques, unbinding the system usage from the need of infrastructures in the surrounding environment.