Development of a Two Factor Authentication for Vehicle Parking Space Control based on Automatic Number Plate Recognition and Radio Frequency Identification

Abstract

This paper proposed a two factor authentication for vehicle access controls using Automatic Number Plate Recognition (ANPR) and Radio Frequency Identification system (RFID) for the University of Zambia (UNZA) vehicle access points. The University of Zambia is experiencing increasing challenge of car parking space and vehicle access controls to and within campus premises. The survey that was conducted reviewed that members of staff found difficulties finding parking spaces due to intrusion. The survey also reviewed that vehicles have been stolen within campus parking areas without detection. An access control system using integrated ANPR and RFID technologies was developed to provide five authentication states that met different vehicle access point's requirement. It was built with 'ORed' and 'ANDed', logic settings to achieve five different states of authentication levels, each suited for a particular access point. The ANRP system used the vehicle number plate to authenticate the vehicle through the use of the camera. On the other hand, the RFID system used the drivers' card/tag through the RFID card reader to authenticate the user. Daily transaction records were sent to the security center where information would easily be retrieved. Illegal access to restricted areas, threats of theft of motor vehicles and failed transaction recording system was amicably solved by this proposal.

A Live-Video Automatic Number Plate Recognition (ANPR) System Using Convolutional Neural Network (CNN) with Data Labelling on an Android Smartphone

“…In auto-gate systems, the ANPR identifies the vehicle license plate number and compares it against a database of registered campus members, and the respective information such as the name of the vehicle owner, student or staff identification number, faculty name and telephone number can be retrieved and displayed automatically. This can be used as a cheaper access control method instead of using access cards or checking campus stickers which is costly, time consuming, and may lead to traffic congestion especially during peak hours [6]. With the aid of Google Cloud Vision, an API that uses machine learning models trained on a huge dataset of pictures that is accessible via API queries, licence plate numbers identified by the system may be captured and logged for future retrieval if the need arises.…”