DRIVER DROWSINESS DETECTION FOR BEHAVIORAL ANALYSIS

Abstract

Driver drowsiness is the leading cause of traffic accidents; thus, drowsiness detection plays an important role in preventing them. By developing an automatic solution for alerting drivers of drowsiness before an accident occurs, the number of traffic accidents could be reduced. As a result, this study suggests a method for detecting drowsiness in real time. Image processing and machine learning are two aspects of the suggested approach. The purpose of the image processing step is to recognize the driver's face and then extract the image of the driver's eyes. This phase employs the Haar face detection method, which takes video as an input and outputs the discovered face. Next, haar is utilized to extract the detective's eye image, which will be used as an input for the machine learning phase. The role of machine learning, in this case, is to classify whether the driver's eyes are closed or open using a support vector machine. If the classification results show that the driver's eye is closed for a predetermined amount of time, the driver's eyes will be regarded closed, and an alarm will be activated to inform the driver. On available benchmark data, the proposed methodology has been tested. The results show that the hybridized image processing approaches with machine learning techniques are accurate and resilient. As a result, it can be concluded that the presented methodology is a viable solution method for detecting driver tiredness in the real world. Keywords- Drowsiness detection: Face detection; Eye detection; Yawn detection; Eye aspect ratio; Haar cascade algorithm.