Image-Based Fatigue Detection of Vehicle Driver: State-of-the-Art and Reference Model

The paper analyzes modern methods of driver fatigue detection. There are a huge variety of methods for assessing the functional states of a person. The detection of the dynamic behavior of the driver in recent years has become an increasingly popular area of research. Dynamic assessment of driver behavior includes long-term monitoring, which allows determining functional states, in contrast to modern driver monitoring systems, which assess conditions such as drowsiness and impaired attention for a short (1-10seconds) time interval. Such systems allow us to talk about physiological monitoring, but not neurophysiological, which allows monitoring the functional state of fatigue. Therefore, it makes sense to monitor the state of fatigue of the driver, as well as to warn him in a timely manner in order to avoid collisions with other vehicles.