Alerting Exhausted Driver using Haar Cascading Algorithm

Abstract

Now a day’s road accidents are significantly increasing with a major reason i.e. due to driver
exhausted or drowsiness which causes the difficulty of focusing and missing the traffic signs. The driver
exhaustion might happen because of continuous and long journey driving, due to these reasons’ drivers
will get exhausted which leads to accidents. So, we need a system that finds the rate of sleepiness of a
driver and inform when he is sleepy or exhausted. To overcome this problem, the proposed system is
introduced, which includes a continuous picture recording device been installed before the driver. The
picture recording system captures the driver’s face and also analyze the movements. The live video of
driver is captured then the image frames are extracted. The unclear RGB images will be amplified [6]
then converted in to gray scale image. The Haar Cascade Calculation (HCC) is used to identify and
detect the inner edges of face, and restores the length, breadth, ordinates, and abscissa of the feature
rectangle-shaped in the face picture particularly the mouth and eyes. The Support Vector Machine
(SVM), then extract the features of drowsiness, depends on the resultant value of the exhaustion discovery
system the driver will get alert from the alerting system.