Drop-Off System Using Machine Learning for Drivers

When fatigued or dwindling, drivers’ natural behaviors, perceptions, and awareness of their surroundings are compromised. Driving at night or under the influence of alcohol is an indication of this lack of caution among drivers, which can result in collisions and pose a serious risk to individuals and the public. As a result, it is critical that the newest developments in the automotive industry include driver aid systems that are capable of detecting driver fatigue and drowsiness. This project provides a nonintrusive prototype computer vision system to track a driver's attention in real time. Eye tracking and groaning are two of the most crucial technologies for future driver assistance systems because allow for the interpretation of an operator's gaze, level of attention to detail, and level of fatigue based on their mouth and eye movements. Convolutional neural networks, commonly referred to as CNNs, are employed in motion recognition and real-time tracking of eyes. A common problem among many of the eye tracking methods proposed so far is that are sensitive to changes in illumination. This typically drastically limits their potential for use in automotive applications. Real-time monitoring and detection of the eye and yawn are a topic of active research in the computer vision field. Movement positioning and eye tracking can be useful for facial alignment. This paper presents an eye tracking and movement detection technique that operates in dynamic, real-world lighting conditions in real time.