Real Time Driver Fatigue Recognition Based on Image Processing

Abstract

Driver fatigue is the main reason of serious damages among all other road accidents. Thus, a new system is proposed with a modern approach which will detect driver fatigue by considering most of the fatigue symptoms such as eye closure, yawning, head tilting. Inattentive vehicle movement on the road under fatigue condition is also considerable for driver fatigue. The approach of this system is to detect these symptoms for the best driving condition on the road. These symptoms are monitored by using two cameras. Thus, a robust system is proposed where head tilting, facial expressions and lane departure for fatigue will be detected collectively. First step of this system is to visualize the face of the driver from one camera while another camera is used to track the road. Then face, eye and mouth (check for yawning) detections of the driver are performed. Facial detection of our method is free from any kind of wrong or bogus detection. Besides, the detection process can strongly determine the desired part of the face in spite of poor light intensity of camera and more realistic to implement. Head tilting detection makes this approach more promising. Lane departure detection in the approach gives this system an advanced level. Lane detection measures the distance of the vehicle between two lanes. Alarm system has been made which will be activated if any fault is found through these detection procedures.