Smart Security and Parking System

Abstract

A video camera will be interfaced with the processor (Raspberry Pi). As the vehicle approaches, the camera will take the image of person and transmit it to Pi for image processing. In the database, images of the persons are already stored for matching process. Face recognition will be used for verification. As soon as the Raspberry recognizes the image, it will signal the Reader to start its own processing. Active tags placed on the vehicles will send the Tag ID related to that particular vehicle to Reader for vehicle identification. There would be three parking areas. The vehicles belonging to a particular class would be directed to go to its particular parking area. Depending upon the ID received, main barrier and the barrier of specified parking will be lifted. For no match case in any of face recognition or RFID techniques, main barrier will not open. For obvious reasons, machine based recognition can never surpass face recognition of a human eye. Thus there remains a room for improvement in terms of efficiency and results as well. Complexity in computation is a major dilemmas of recognition techniques and for large databases, this problem is further aggravated. In this project the algorithm used is Principal Component Analysis with the help of which face recognition process will be achieved in an easy and low computational manner.