Drowsiness Detection and Safety monitoring using Image Processing

Drowsiness Detection and Safety monitoring using Image Processing

Drowsy driving is a major issue of the traffic collision due to which severe injuries and deaths occur by the way of accidents. Numerous researches were undertaken to carefully design systems which can assess fatigue of the driver and provide an alert prior, hence preventing from sleep avoiding accident. Some conventional approaches employed vehicular measures that are greatly influenced by the road geometry, vehicle model, driving skill etc., behavioral measures like facial expression and psychological subjective measurements that can cause discomfort to the driver and may produce ambiguous results. In this paper, a nonintrusive and real-time approach is proposed that monitors driver’s eyes using a high resolution Pi camera. An algorithm is developed to capture the symptoms of driver fatigue through facial and head movements. On detecting abnormal conditions, a buzzer alerts the driver and an alert message will be sent. Raspberry-Pi is used to incorporate the entire system. Experimental outcomes ensures that the system could track the changes in the driver’s facial movements and alert the driver from accidents more efficiently with an appreciable accuracy. The system can be implemented at any lighting condition.