Artifact correction and refined metrics for EEG-based fatigue detection system
Description
Hamburg, November 2019.The aim of the work presented at the Congress of the German Sleep Society is to recognize the driver's fatigue. EEG signals are divided into sequences by an algorithm and classified by a neural network.
Prof. Dr. Martínez Madrid during the presentation in Hamburg.
Fatigue is often underestimated, but a still big problem in road traffic. Out of about 2.5 million road accidents in Germany in 2015, 2,898 accidents, with a total of 59 fatalities (~1.7% of deaths), were due to fatigue. Estimates assume a number of unreported cases of up to 20%. On the basis of the Portable System to Detect Driver Drowsiness with Body Sensors (PoSDBoS), an electroencephalogram (EEG) is used to detect fatigue. The aim of the work is to improve the measurement system used. The accuracy is increased by artifact correction and refined quality metrics. Detected fatigue is then appropriately displayed to the driver so that he can react accordingly.