Driver Drowsiness Detection Systems

Car accidents are one of the biggest concern in society and driver drowsiness is also a significant factor in a large number of car accidents. To improve public safety and the reduction of accidents, There have been different developed approaches and methods to tackle this problem. The aim of this story is presenting these methods and how they can detect driver drowsiness. In the end, Advantages and limitations of methods are mentioned. All methods of driver drowsiness detection are shown in figure 1.

Figure 1
Figure 1

Drowsiness detection methods are separated into two main categories: methods focusing on driver’s performance and methods focusing on the driver’s state.

Driver’s Performance-Based Methods

For detecting drowsiness, studies on driver’s performance employ lane tracking, the distance between driver’s vehicle and the vehicle in front of it; place sensors on components of the vehicle such as steering wheel, gas pedal and analyze the data taken by these sensors. Pilutti and Ulsoy used vehicle lateral position as the input and steering wheel position as the output and they obtained a model which can be useful to detect drowsiness. Some of the previous studies have used driver steering wheel movements and steering grips as an indicator to detect drowsiness. Some car companies such as Nissan and Renault used have focused on driver performance method to detect drowsiness. Since this method is too dependent on the characteristics of the road, they can only function well on motorways which make them work in limited situations. The first limitation of this method is that it is affected too much by the road quality and lighting. Another disadvantage of this method is that it cannot detect drowsiness that has not affected the vehicle’s situation yet. It means when a driver is drowsy and the vehicle is in the appropriate lines, these systems cannot detect drowsiness.

One of driver drowsiness detection system based on steering wheel

Driver’s state-Based Methods

Methods focusing on driver’s state are separated into two main groups: methods using Physiological signals and methods using driving behaviour.

Methods Using Physiological Signals

The methods employes physiological signals such as Electrooculography (EOG) Electroencephalography(EEG), heart rate variability (HRV), pulse rate and breathing. The spectral analysis of heart rate variability shows that HRV has three frequency bands: high-frequency (HF)band (0.15–0.4 Hz), a low-frequency (LF) band (0.04–0.15 Hz) and very-low-frequency band (0.0033–0.04Hz). Researchers have found out that the LF/HF ratio decreases and HF power increase when a person goes from alert state to drowsy state. The power spectrum of EEG brain waves is used as an indicator to detect drowsiness; as drowsiness level increases, EEG power of the alpha and theta bands increases and beta band decreases. However, EEG-based drowsiness detection methods are not easily implementable because they require the driver to wear an EEG cap during driving the vehicle. Devices being destructive is the main disadvantage of this group of methods.

Automated Detection of Driver Drowsiness with EEG signals

Driving Behaviour Detection-Based Method

This group of methods are not offensive and does not make any disturbance to the driver, that’s why these methods are more preferable. These methods can be divided into two groups: the ones using infrared illumination and the ones using day illumination. These techniques involve computer vision systems that can detect and recognize different facial expressions and the changes in facial appearance that occur during fatigue states. This method is also non-invasive and has the advantage of using computer vision techniques, and therefore being more amenable to use by the general public. Computer vision techniques employed to detect fatigue have focused on the analysis of blinks, head movements and facial expression such as mouth, yawning, and nose wrinkles.

This camera can detect the driver’s face and also detect facial expression.

The summary of the Advantages and Limitations of Driver Drowsiness Detection Systems (DDDS)

I am a master student in Human-Computer Interaction. I am working on designing the interaction between Human and Autonomous Vehicles, Driver Fatigue Detection.