Certain investigations on drowsiness detection system using facial features and heart rate

Abstract

Driving a vehicle with utmost safety measures is always required to prevent accidents at the wheel Most of the accidents that occur in every part of the country are due to driver fainting at the wheel Major causes for accidents interms of loss of attention and diseases that may inhibit with the benign operation of vehicles can be found in the following factors such as least correct vision Earshot problems Cardiac diseases Cerebral Stroke Effects of using sedatives and aging issues that provides the decrement in strength slow reaction time lack of responsiveness and misperception Among these issues fatigue and sudden heart attack are the major causes of accidents which cannot be controlled by driver even if he wants to do It is important to point out the distinction between prediction and detection of drowsiness Obviously prediction in advance is the main task since at the detection point drowsy driving may already have led to a potentially hazardous situation or even an accident. Initial step starts with the acquisition of an image from the camera followed by face tracking eye detection mouth detection and drowsiness detection which is described in detail below For performing the face detection the first step is pre processing It is used to calculate the integral images and lighting corrections from the scaled original image The second step is processing the frame sequences along with trained threshold values The third step is post processing It clusters the adjacent detected faces to one or several rectangles to represent faces according to distance factor From the resultant binary image collect the number of connected components and check if there are two connected components left to find the eye blink If so then an eye blink has been occurred newline