Smartphone Based Driver Aided System to Reduce Accidents Using OpenCV

Abstract: Road traffic is increasing rapidly because of availability of various public and private means of transportation. Flow of traffic rarely stops at any hour of the day because of extensive work schedules and travelling needs. This ultimately leads to long driving hours by not only people who are driving continuously to face the adverse effect of fatigue i.e. drowsiness and sleep deprivation. Driver aided system uses android based smartphone which helps to detect driver under fatigue and system alerts driver under sleepy conditions. It is real time driver fatigue detection system which uses OpenCV (Open Source Computer Vision) for tracking driver�s facial expression for preventing accident. Proposed algorithm tracks eyes, mouth to detect doziness and yawning respectively. It also detects stress from driver�s facial expressions like forehead lines and give warning to passengers. It also provide stress recognition through driver�s speech like driver asking for help and takes respective action. Warning system uses different levels of warning includes messages, voice messages, beep, vibrations to alert driver. This system HaarCascade filter libraries for facial tracking and other face region tracking.

Keywords: Drowsiness, OpenCV, HaarCascade, facial tracking, smartphone.