Drowsiness Detection Model Using Machine Learning Techniques

Abstract

Road accidents are a major hazard to public safety around the world, killing many people and generating considerable economic and social consequences. These incidents are caused by a variety of circumstances, including human mistake, such as driving under the influence of alcohol or drugs, distraction, and exhaustion. Driver sleepiness is one of the most dangerous conditions on the road. When drivers grow tired, their ability to concentrate and respond rapidly to changing road circumstances suffers dramatically. Studies have indicated that drowsy driving can impair cognitive skills, resulting in longer reaction times, worse judgment, and diminished awareness of one's surroundings. As a result, fatigued drivers are more likely to make important errors, such as straying out of their lane or failing to spot dangers, which increases the risk of an accident. To reduce the risks of drowsy driving, advanced driver assistance systems (ADAS) have been created, with drowsiness detection systems playing an important role. These systems use technical breakthroughs, particularly in computer vision and machine learning, to monitor driver behavior and detect indicators of drowsiness in real time. The use of machine learning techniques, such as the Eye Aspect Ratio (EAR) and Histogram of Oriented Gradients (HOG), has showed potential in detecting drowsiness. The EAR algorithm examines the ratio of ocular landmarks, such as the distance between the eye's corners and its height, to detect drowsiness-related alterations, such as drooping eyelids or prolonged periods of eye closure. Meanwhile, the HOG algorithm gathers characteristics from the driver's face, such as gradients and edge orientations, to detect small changes in facial expressions and movements that indicate tiredness. Drowsiness detection systems can effectively measure the driver's level of attentiveness and act when necessary by combining machine learning techniques with data acquired from onboard cameras. For example, if the system detects signs of tiredness, it might inform the driver with auditory warnings or seat vibrations, pushing them to fix the situation or pull over to rest. These devices can also provide useful insights about driver behavior patterns and fatigue trends, allowing for preventative measures to reduce sleepy driving incidents. Fleet operators, for example, might use data from sleepiness detection systems to apply fatigue management tactics including arranging regular breaks and adjusting work schedules to prevent driver weariness.