A survey on visual and non-visual features in Driver’s drowsiness detection

Many road accidents are happening due to the negligent behaviour of the drivers, which increases the death rate day by day. The tiredness and drowsiness of the drivers are the primary cause of road accidents. Due to technological advancement, various techniques evolved to identify the drowsy state and alert the driver. As per the literature, the drowsiness detection techniques are categorized into three classes based on driving pattern, physiological characteristics and Computer vision. Among these techniques, we have focussed mainly on the Computer Vision technique in our survey due to its low cost and non-intrusive nature. This technique analyses the various images of driver’s posture, such as facial expression, yawning duration, head movement and eye closure to identify drowsy state. A detailed comparative study is presented in this paper and observed that spatial feature based techniques have given highest result with precision 97.12%. Also, state-of-the-art drowsiness detection techniques are exposed, analyzed and reviewed rigorously.