Driving fatigue is a serious issue for traffic accident risk, for fatigue affects the drivers' ability to continue driving safely. On-line monitoring of physiological signals while driving provides the possibility of detecting fatigue in real time and is absorbing much attention in the road safety field. This study aims to identify driving fatigue through a driver's physiological signals. Physiological signals, including heart rate, skin conduction, electro- myogram, skin temperature and respiration, were collected on eight simulating experiments with 1Hz sample rate. Then wavelet transform was used to find feature vectors and a two-step fuzzy cluster analysis was developed to classify the different levels of fatigue, including alertness, a transitional phase and fatigue. Compared with the experimental data, 93.75% of the normal and abnormal states and 63.64% of the transitional and fatigue levels had been correctly identified. The results showed that such a method could provide an effective way to detect a driver's alert state level. To improve the reliability of this algorithm, more physiological signals are needed and an impersonal metrics must be used to judge the actual state of the driver.
Driving Fatigue Identification Method Based on Physiological Signals
Seventh International Conference of Chinese Transportation Professionals Congress (ICCTP) ; 2007 ; Shanghai, China
2008-03-21
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Classification of driving fatigue based on EEG signals
IEEE | 2020
|Fatigue indicators of drowsy drivers based on analysis of physiological signals
Tema Archiv | 2001
|A Driving Fatigue Identification Method Based on HMM
SAE Technical Papers | 2020
|A Driving Fatigue Identification Method Based on HMM
British Library Conference Proceedings | 2020
|