Drivers Fatigue Detection and Alarming Using Ear and MOR Values

Abstract

There are many accidents that occur daily on highways, roads, and there are many reasons for it to occur. According to surveys done the major cause of the accident was either the driver was not completely awake or they were feeling drowsy and this resulted in accidents either major or minor. Hence, detection of the driver’s fatigue and alerting on time is the main motive of this research. Few methods which detect drowsiness are intrusive and distract the driver whereas some methods of heavy sensors are installed which are expensive. Therefore, in this study, a budget-friendly, a real- time driver’s fatigue detection system is developed. The proposed system incorporates the Histogram of Oriented Gradients [HOG] algorithm and Support Vector Machine [SVM] algorithm for face detection and Regression Trees algorithm to get the facial landmarks and calculate two major factors which can help us determine the status of driver fatigue. The two major factors that help us determine the driver's fatigue are eye aspect ratio, mouth aspect ratio. Here, this research study will calculate the values first when the driver of the vehicle is completely awake. Such values will be recorded and later it will be set as threshold values. After threshold values are set, the Eye Aspect Ratio and Mouth Opening Ratio are recorded. Then the recorded values are compared with the threshold values. If there are deviations in the values measured then the system sends out an alarm saying that drowsiness is detected. This proposed system helps in the reduction of accidents happening due to Drivers fatigue.