Real Time Rear Vehicle Monitoring & Detection System

Abstract

The purpose of this project is to design and develop real-time vision based systems which can identity vehicles approaching from the rear and alert the driver in order to prevent a possible collision. The camera placed on the rear side of the vehicle, which captures images at a rate of 30 frames per second and then sends it to the processor via transceiver systems such as Wi-Fi. These images are analysed by various images processing and video processing techniques, the processor finds whether the vehicle behind is in a danger position or not. If a danger is sensed then an alert is given to the driver in the form of beep or voice based alerting system. A beep sound is given before a dangerous situation arrives, so that the driver is mentally prepared so do the right manoeuvres. The proposed system will also guide drivers while taking turns, if suppose the driver wants to take a left turn, the area of interest of camera will concentrate on the left portion and if the driver wants to take a right turn the camera area of interest shifts towards the right portion.