Drowsiness Detection System with Speed Limit Recommendation using Sentiment Analysis

Abstract

Driving while drowsy is a ubiquitous and extremely grave public health hazard that requires immediate consideration. Through studies in recent years, it has been proved that about 20 percent of all car accidents have occurred as a result of dozy driving. The main objective of new drowsiness detection systems is accurate doziness recognition. In this regard, the face is the most important part of the body as it sends a lot of essential information. The facial expressions of a drowsy driver include frequency of blinking and yawning. This paper proposes a model which detects the drivers’ awareness using video stills of the driver’s face and improves the tracking accuracy. Further, we introduce the auxiliary functionality of speed limit recommendations based on the driver’s present state of mind. The various facial features are evaluated to determine the drivers’ current state. By combining the features of the eyes and mouth, the driver is alerted with a fatigue warning and also suggested a safe speed limit. This system is very essential so as to prevent and hence reduce the number of fatal accidents that occur as a result of dozy driving saving a lot of lives and damage to property.