Driver Drowsiness Detection In Real Time Using Deep Learning and OpenCV
University of Texas at Arlington Real-Life Drowsiness Dataset
There have been several studies and contributions concerning the safety and security of vehicles. Road accidents have emerged among the pronounced causes of death. However, in recent years, an unconscious error from the driver’s bit has surfaced and has been proven as a critical factor for causing road accidents. Drowsiness has causes like sleep deprivation, tiring journey, and alcohol consumption. Drowsiness slows the driver’s reaction time to be able to apply brakes or apply a sudden steer and influences the driver’s judgment capability. In situations like these, an individual is not in com- plete supervision of his own body. Once it is clear that the driver is not active, all the solutions yield only one question, whether the driver has to be alerted or the vehicle should be stopped or powered down. Since then, driver drowsiness and its reasons are being investigated to provide a real-time, practical solution.There are several solutions suggested that provide assistance that alerts the driver or takes control of the vehicle if drowsiness is detected.
This project aims to consider facial features like the eye and the mouth and study these facial features to predict using Open CV to know if the driver is conscious enough. It proposes Deep Learning-based algorithms on the calculated eye closure and mouth distances, yawns to detect drowsiness.
