Accident Prevention Using Drowsiness Detection

Accident prevention is a vital aspect of ensuring road safety, as road accidents can lead to severe injuries and fatalities. Therefore, effective measures to prevent accidents are of utmost importance. One prominent cause of car accidents is driver drowsiness, which impairs the driver's ability to react promptly, increases reaction time, and reduces alertness. To address this issue and minimize the risk of drowsiness-related accidents, we propose a novel project that employs driver drowsiness detection techniques based on eye detection, specifically the Eye Aspect Ratio (EAR), as well as lip detection, utilizing the Lip Aspect Ratio (LAR). Our proposed system actively monitors the driver's eye movement, blink rate, and lip movements in real-time to detect signs of drowsiness. The system utilizes a camera mounted on the car's dashboard to capture the driver's face, and the collected footage is transmitted to an analysis computer for processing. Computer vision techniques are then employed to detect and track the driver's eyes, calculating the EAR, which measures the ratio of the eye's height to width. Additionally, the system identifies the driver's lips and calculates the LAR, a ratio representing the mouth's height to width. To build the suggested solution, deep learning algorithms have been extensively trained on extensive datasets of images and videos of drivers exhibiting various levels of intoxication and sleepiness. These models have been taught to recognize distinct signs of sleepiness, such as drooping eyelids, sluggish eye movements, and frequent yawning. When the driver's sleepiness levels exceed a predefined threshold, the technology alerts the driver and the car's control system. These alerts can take the form of visual or audible indications, guiding the driver to stop, pull over, or take other necessary safety precautions. The suggested system offers several potential advantages, including improved traffic safety, reduced frequency of drowsy-driving-induced crashes, and decreased accident severity. The valuable insights into driving behavior obtained through this technology can drive advancements in systems designed to cater to the needs of drivers, benefiting automakers and organizations responsible for road safety. Moreover, the proposed technique holds promise for application in various other domains, such as aviation, where fatigue and drowsiness pose significant concerns. In conclusion, our project aims to develop a real-time driver sleepiness detection system using LAR, EAR, lip, and eye detection algorithms. By implementing this device, we anticipate a substantial increase in traffic safety and a reduction in accidents caused by drowsy driving.