Driver’s inattention and distraction are the most prominent reasons for road accidents. Since a fraction of second distraction may cause a severe accident and hence active attention of driver is mandatory while driving a car. Intelligent driving assistance system may reduce accident rate that are mostly occur due to inattentiveness and in turn improve the efficiency in driving. It is quite challenging task for computer vision to monitor the driver’s level of attention continuously and assist him/her when level of attention is low due to distractions. This paper presents a driving assistance system that can compute the driver’s attention and determine his/her level of attention while driving using a simple webcam and computer vision technique. The driver’s attention level is determined by estimating his/her face direction, gaze direction, mouth movement, and head pose from video stream captured by the camera. If the level of attention crosses a predetermined value, the system initiates an audio sound to create the alertness of the driver. Experimental results show the system is functioning well and successful to generate alarms for 89.34% cases of inattention.
- [OpenCV] (https://github.com/opencv/opencv)
- Dlib
- Deepgaze
- Python
- Java