文摘
In this paper, we proposed a method that classifies electroencephalography (EEG) from color imagination data using the Emotiv EPOC headset. For EEG measurement and the event-related potential (ERP) method, brain-computer interface (BCI) systems were used in the experiment. In the experiment, the subjects gaze at a non-flicker visual stimulus of color (i.e., red, green, blue, white, and yellow) and then proceed to imagine the color. To concentrate on the LED light, all experiments were performed in a dimly lit room. The flickered visual stimulus was made using an Arduino microcontroller board and LEDs with the purpose of prompting color imagination. As a result, we obtained significant EEG responses of thoughts related to certain colors. The EEG response is classified using classification algorithms including a support vector machine (SVM) with linear discriminant analysis (LDA), an artificial neural network (ANN) with LDA, and an ANN without LDA. In addition, five-fold cross validation was used to evaluate the performance. From the results, we found robust electrodes (T7 and F4). The technology developed in this paper can be used to assist paralyzed individuals and the elderly.