Detección del estado de somnolencia para conductores mediante el análisis de ondas cerebrales

Abstract

To develop the drowsiness detection system using brain waves, electroencephalographic data of drivers in driving states with and without drowsiness are collected based on established traffic and time conditions, from the recorded recordings the frequency band data is accessed and visual patterns of the behavior of the Theta, Alpha and Beta brain waves that differentiate the state of sleepiness compared to the active state of conduction are sought. Additionally, sleep recordings with sleepiness were taken that helped in identifying sleepiness patterns. Python programming accessed and processed the frequency band data for Theta, Alpha, Beta and AF3, AF4 electrodes located in the prefrontal area of the skull, identifying ranges of amplitude in the frequency band that classify a conduction with and without sleepiness, the time that the data remains within the corresponding range is also determined. It met in a Python-based graphical interface, processes such as the initial connection with the headband through Websocket, the capture, data recording and analysis of the program in real time, as well as the activation of the auditory and sensory actuators that work as an alert to the driver. Validating the values and conditions implemented for the detection system, performance tests were conducted on drivers in the driving state with and without drowsiness, verifying that the time and permanence of the data in the amplitude window has been fulfilled.