One of the paradigms that has been mostly used in the literature for the implementation of an EEG BCI system is the visual evoked potentials in steady state, which normally arise in the occipital cortex of the brain. There is a series of stages that are required in order to visualize, extract, and classify them. The initial phases of the methodology in this study were divided into acquisition, pre-processing, extraction, and classification, just as in the design of an IBD system. This study makes a characterization of these potentials, from the acquisition using the g. Nauti-lus equipment with the 10-20 standard of Universidad Antonio Nariño until the classification of the data using the CCA and SED mathematical methods in different time windows. Thus, the implementation of a realtime BCI system is supposed to have a classification time that is as short as possible for the rapid execution of a command; this type of study allows the identification of the best methods to be used in the classification of these data, as well as some variables that can be taken into account. The results allow to identify, then, greater effectiveness in the classifica-tion of data with CCA than with SED, as well as the fact that the system behaves according to the time window.
