Applying HRV Based Online Clustering Method to Identify Driver Drowsiness

Abstract

Traffic accidents related to driver drowsiness are a major safety issue. Many research studies propose new methods to reduce the number of drowsiness-related injuries and fatalities. The ultimate goal for a drowsiness detection system is to detect the drowsiness on time and minimize the system or environment errors to avoid false readings, such as studying physiological signal processing patterns. Electrocardiogram (ECG) research studies the change of the heart rate variation signal related to human psychological behavior. In these studies, measuring heart rate variability (HRV) parameters in the frequency domain formed from the periodic ECG is proven to correlate with drowsiness levels. In this paper, we apply unsupervised machine learning (clustering) to study the behavior of HRV during drowsiness. With this method, we can measure different levels of drowsiness based on the changes in the density and shape of the HRV clusters. Moreover, in this method, the pre-measured labeled data is not required to establish the algorithm. Therefore, for any unknown object or person, this algorithm evaluates drowsiness, and no prior data is required.