Real-Time Drowsiness Detection Using Fusion of Facial Features

Drowsy driving is one of the biggest causes of road accidents and deaths. The proposed methodology explains a state-of-the-art examination based on recent computer vision-based drowsiness detection techniques. Driver weariness results in a lack of concentration, which directly contributes to a higher rate of accidents. The dilemma highlights the need for a system that detects the driver’s tiredness and provides an alert signal to the driver prior to the accident. The work proposes a method to detect negligent behavior or drowsiness based on facial features. The system uses a webcam to record the video, and image processing algorithms are employed for the extracted frames. The eye aspect ratio and mouth opening ratio are computed for each frame utilizing the facial landmarks followed by adaptive thresholding. The system alerts the driver if drowsiness is detected. The technique is effective in cases when the drivers are accustomed to a demanding workload and are required to drive constantly for long periods of time.