Site icon Visionify

How to detect driver drowsiness and send alerts?

Veoneer Driver Monitoring System

Driver drowsiness can be a safety hazard, especially on a long trip. People often do not realize they are too tired to operate a motor vehicle until it is too late. It may become evident when your eyelids start drooping, and you have trouble keeping your head up. You may even realize that driving is so effortful that you cannot hold the steering wheel steady. Driving while drowsy or fatigued is associated with an increased crash risk and adverse effects on driving performance. Many crash-involved drivers report having fallen asleep or otherwise become unaware of their surroundings (e.g., “zoning out”) just before the crash. Modern cars with driver drowsiness detection help to avert collisions. In this blog, you can see how drowsiness is detected and send alerts.

Markets and markets say the Global ADAS market size is projected to grow from USD 27.2 billion in 2021 to USD 74.9 billion by 2030.

Driver drowsiness detection app

The driver drowsiness detection app can save car drivers by identifying fatigue in motorists driving habits. The app uses a light meter and sound meter to measure a person’s state of the nervous system. It can infer the person’s mental status with these measures. 

Stress and tiredness markers are measured. Then mental status is measured quantitatively as a dynamic stress index (DSI). DSI ranges from 0 – 100. 

A lower DSI value indicates a higher level of concentration. Driver drowsiness detection app shows an alert when DSI falls below the selected level. It suggests stopping driving when the DSI level is below the critical level. It also estimates braking distance with considering factors. 

Drowsiness detection and alerting system

Drowsiness detection and alerting system (DDAS) is a computer application developed to check if the driver falls asleep. When he is getting drowsy while driving. The system detects that the driver is dozing off, frequent yawns, prolonged eyelids closure, and reduced subtended angle of the eyes.

Video image processing, visual pattern recognition, and alarm devices are all used to detect driver drowsiness. The system configures the results of the recognition unit to set off an alarm. The visual pattern recognition unit compares the real-time video. 

The previous video determines whether or not the driver was sleeping. Then the alarm unit outputs the alarm according to the comparison result of the recognition unit.

DDAS detects drowsiness alerts the driver when getting into a dangerous situation. The main objective is to prevent road accidents as drivers’ impairment causes them sleepiness. Based on monitoring the driver’s physiological signals.

As an example (thermoregulatory, pulse, blood pressure, skin conductance, and EEG-based). In addition, medical sensors installed in the car regularly collect data.

Driver drowsiness detection dataset

The system takes advantage of the computer vision driver drowsiness video dataset. The entire data set includes various peoples, testing data, evaluation, and training. 

Measurements were obtained when not wearing sunglasses and when wearing sunglasses. Included a range of driving scenarios such as laughing, falling asleep, slow blinking, yawning, and typical driving as well as night and day illuminations. The subjects were driving an automobile, directing the wheel, braking, and changing their facial expressions. All sequences define drowsiness and non-drowsiness. Analyze and keep in a dataset for future applications and research.

Driver drowsiness detection using CNN

Driver drowsiness detection system is a safety-critical application. It requires careful design, development, and testing. Driver state inference from vehicle data has many challenges. Include detecting drowsiness in real-time despite possible obstructions such as sunglasses and car windows. 

We can see the convolutional neural network (CNN) built for driver drowsiness detection. Based on image recognition technology

first of all, it is necessary to represent the imaging data space of the driver’s face in a concise format.

Then sample images are provided for training and validation of CNN for better accuracy. The results show that driver drowsiness detection systems can operate effectively with a high accuracy rate. Under different lighting environments, we are wearing colored glasses.

A CNN comprises multiple layers, each extracting more meaning from the input. For example, CNN divides images into regions for driver drowsiness detection and selects an appropriate part depending on their location. Next, evaluate their suitability as eyes; moreover, it estimates the driver’s state by analyzing their eyes. As a result, the drowsiness deters quickly.

Source: Youtube

Conclusion

We need to be alert and attentive to avoid accidents. People who go drowsily are a vast security threat to themselves and others. We should develop applications to detect driver drowsiness and send alerts on behalf of those who cannot do so.

Visionify provides many computer vision solutions and various ways to help your business effectively. We collaborate with clients to develop solutions tailored to their specific requirements. We have worked extensively on solutions in the manufacturing, retail and food industries. Call us to get a live demo.

Exit mobile version