Real Time Drowsiness Detection System

With the advancing technology in digital computing world, it has become a lot easier to detect the very minute details of everything. This work is intended to detect the state of drowsiness in drivers in order to minimize the accidents and to improve road safety. The implementation of real time drowsiness detection is done by locating driver’s facial gestures along with eye parameters. Our effort is to detect the eye blink of the driver. If the eyes of the driver are in closing state for more than the threshold time, it will show a drowsiness alert and arouse the alarm. The coding is done in python language and Open CV is added for detecting facial features. We aim to develop a real-time algorithm to detect blinking of eyes from a video stream captured from a standard camera. The precise detection of landmarks was enough to estimate the opening of eyes. We propose an algorithm which estimates the landmark positions by extracting a single entity – eye aspect ratio (EAR) – and characterize the opening of eye in each frame. Finally, a Support Vector Machine classifier detects eye blinks as a pattern of EAR values in a short temporal window. The simple algorithm outperforms the state-of-the-art results on two standard datasets. General Terms: Facial Recognition, Landmark Localization, SVM Classifier