Realization of VLSI architecture to detect driver drowsiness for road accident avoidance system

Automobile Industry shares the massive distinction of being part of numerous accidents in our daily routine. The main objective of this work is to develop a vision zero approach that promotes a high standard integrated safety culture and achieve a secure environment with no fatalities for prioritization of road safety. Increasing rate of road accidents are due to driver distraction such as fatigue and lack of sleep. Thus, the important fact to be considered is the prevention of sleep induced road accidents. In this work the VLSI architecture based classification is used to define the subject as drowsy or non-drowsy. In the proposed system, the image is preprocessed using median filtering. The filtered image is subjected to cascade object detection for extracting the facial features. Further, the features are extracted by using Local Binary Pattern analysis and the features are reduced .Then the Gaussian kernel unit as well as decision module is used to define the status of the subject as drowsy or not. Gaussian kernel unit involves in computing the normal processing element as well as exponent processing element. Cordic exponent PE is used in order to reduce the area utilization. This system is developed for warning the drivers to prevent accidents.