Main Article Content

Development of IoT-based smart door unlocking in cars using a hybrid of face recognition and personal information number login system


Aderonke J. Ikuomola
Oluwatosin E. Bashir

Abstract

Access to most automobiles utilizes key and lock systems. Some newer automobiles make use of a Remote Keyless System (RKS). The crucial drawback of key or keyless system is that these systems grant access to whoever holds and applies the key or remote device mostly without any form of user authentication which creates room for unauthorized access and car thefts. In this work, an Internet-of-Things based (IoT-based) smart door unlocking in car using face recognition and personal identification number (PIN) login was developed. This system comprises of a camera sensor, IoT-reporting module, face recognition engine, car central lock, and authentication login module. The system was implemented using Javascript programming language, Face-API library, Chrome browser, HTML5 Canvas element, and PHP programming language, and was tested within a simulated environment. The simulated system utilizes the Face-API Javascript library, which is built on the TensorFlow JS machine learning framework for face recognition. The result shows an accuracy of87%, which is within the accuracy range of reliability of most face recognition systems.


Journal Identifiers


eISSN: 2636-6134