Efficient eye states detection in real-time for drowsy driving monitoring system

In this paper, we propose a reliable method of eye states detection for drowsy driving monitoring. Given a restricted local block of eye regions, the Local Binary Pattern (LBP) histogram of the block is extracted and each bin of the histogram is treated as a feature of the eye. An AdaBoost based cascaded classifier is then trained to select the significant features from the large feature sets and classify the eye states as open or closed. According to the states of the eye, the PERCLOS score is measured in real time to decide whether the driver is at drowsy state or not. Experimental results demonstrate that our eye states detection algorithm can give an average eye states detection rate of over 98% under different illuminations, face orientations and subjects. This reveals that our method can work reasonably well for the purpose of driver awareness.