Vehicle Security System with Obstacle and Drowsiness Detection

Abstract

In this paper, we aim to design a vehicle security system providing safety-critical systems. The aim is to make a vehicle security and driver assistance system. We try to install a face detection system for the vehicle to start. Only if the authorized user starts the car, it will start. The face detection system works by using raspberry pi and image processing concept. Earlier, an alarm system was used for protection which was not effective. In such systems, it is easier for the car thief to deactivate the security system. In case the alarm goes off, nobody pays attention to it. We also try to develop an “obstacle detection system for vehicles” capable of reducing the increasing number of vehicular accidents on highways. Driver should always be alert to react for the obstacles in way in order to prevent accidents. Using ultrasonic sensor, our system is capable of alerting the driver and stops the motor when the distance between the obstacle and the vehicle is lesser than 10 cm, thus reducing the chance of an accident. Often the tiredness of driver leads to many serious road accidents. Therefore, we are developing a system that will detect a drowsy driver and alert the driver to reduce the number of fatigue-related car accidents. This is also achieved by using face detection. We use LBP (Linear Binary Pattern) for face detection and haar cascades for feature extraction. We try to create an efficient accident alert system that will send an alert message to rescue team when accident is detected thus saving lives without wasting time. Whenever, there is a tilt in the axis of the car, it is easily detected using an MEMS accelerometer, and a message sent immediately to the registered mobile numbers using the GSM module.