Improved Method for Driver Drowsiness Detection Using Hybrid Template Matching Algorithm

Now days car accidents because of driver distraction and fatigues are causing the many damages to human being. Therefore, recently different methods introduced by many researchers for early detection of driver drowsiness in order to prevent accidents on road. The methods are commonly known as driver face monitoring methods introduced for driver fatigue detection as well as accident avoidance. Most of methods are based on image processing concepts in which images of driver face are captured, and then extracted the eyes, mouth and head position of driver from input image. The symptoms for detecting the drowsiness are eyelid distance, eyelid closure over time, yawing, head orientation, head nodding, yawning etc. In this paper our goal is to present novel framework for detecting the driver drowsiness or fatigue efficiently using the Viola Jones Algorithm for detecting mouth, eyes, nose detection and face detection along with template matching approach. Viola Jones method helps to find face, left eyes, right eyes, nose, and mouth points on input image. Local features are extracted and used for template matching process to detect whether mouth is yawing or whether eyes are open or closed. Most of existing methods uses either yawing technique of eye blinking technique for driver drowsiness detection, in this paper we proposed hybrid method which is combination of template matching of open eyes and closed and template of yawing.