Automatic Electrooculogram Classification for Microcontroller Based Interface Design

In this paper, we present a simple and novel technique for classification of multiple channel Electrooculogram signals (EOG). In particular, a viable real time EOG signal classifier for microcontrollers is proposed. The classifier is based on Deterministic Finite Automata (DFA). The system is capable of classifying sixteen different EOG signals. The viability of the system was tested by performing online experiments with able bodied subjects.