EEG Signal Processing Model for Eye Blink Detection

: Electroencephalography devices such as the OpenBCI Cyton Biosensing board create a noninvasive and inexpensive way of acquiring signals generated by the brain. These signals are influenced by different types of brain stimuli such as eye blinks but they are also includes a large amount of noise, e.g., generated by the board. However, the noise can be removed with the help of proven filters. In this aspect, the intention of this work is to demonstrate how using different type of filters, it is possible to clean the noise from the brain signals acquired using an encephalography devices (such as Cytonbiosensing board) which are generated when a user blinks his/her eyes and classify them in different type of blinks. We have chosen the study of eye blink brain signals, since, they present a wide range of real-life applications. Our model includes a simple algorithm that classifies user-generated eye blinks into short intended blinks and long composed blinks. Experimental results of the proposed model show an accuracy of 96% which enables the development of real-life applications that do not require real-time control such as IoT devices.