In this paper we present an electronic circuit
capable of acquiring Electrooculogrphy (EOG) signals. We apply
the Common Spatial Pattern (CSP) and Linear Discriminant
Analysis (LDA) methods to perform the classification of such
signals. The experiments and tests for the creation of a dataset
composed of EOG data acquired from six different subjects is also
described. Four classes, Look Left (LL), Look Right (LR), Look
Up (LU) and Look Down (LD) were classified and a performance
above 95 % was obtained in all the experiments.