Automatic Driver Awake Alert System

Abstract

Automatic sleep detection system has become a most vital system in the 21st century. It should be implemented in all road vehicles as drowsy driving is an invisible killer for drivers if they are driving on highways. Moreover, it is undeniably quite hazardous for traffic security. In real time, catching drivers sleeping behaviour should be implemented to detect the driver fatigue in the early stage itself to prevent accidents. Most vehicle manufactures are full of activity in designing drowsy-driving monitor devices. Most of them have deployed advanced test technology combining embedded systems and intelligent control technology to design their drowsy driving monitor systems. Although these technologies reduce the traffic accidents caused by drowsy driving, they are quite expensive and inconvenient to the driver. Moreover, they are not available in country like Sri Lanka. This Automatic Driver Awake Alert System (ADAAS) is also intended to solve the above critical problem. Unlike other approaches, here the solution was developed using the area of computer vision and Image processing techniques. As a scientific discipline, computer vision is concerned with the theory for building artificial systems that obtain information from images. Implementing this system with the use of computer vision and digital image processing techniques is beneficial. It also eliminates the drawbacks of earlier system as its more accurate, reliable, economical and convenient for users. Furthermore, this technology suits for developing countries like Sri Lanka. This system was developed with the use of a webcam. Webcam generates video files of driver and then they are processed by ADAAS. This system continuously monitors the driver s eye to see if the eyes are closed or open and make noise when the driver gets to sleeping mode. It assumes that driver is sleeping if the eyes are closed continuously for some times. The solution is implemented using the images obtained from the monitoring system. Moreover, these images are continuously preprocessed and analyzed to find the difference with the original preprocessed image (Image of eyes which is decided by the system that the driver is awaken). Two major researches were carried out as part of this project; face detection and eye detection in the frontal facial images. Most researches are related to Computer vision and Image Processing techniques. Having these researches also in the same area was helped to get a concrete solution for this context. This system consists of four components Video acquisition, Face detection, Eye detection and Sleepy analyzer. Carrying out continuous analyses with these components leads to a perfect outcome. This project would be more demandable and usable for people who are drivers as well as who get sleep when they are doing serious works. Key words: Image Processing, Machine Vision, Fatigue, Visual System, Face Recognition.