Prevention of unauthorized door access using face recognition built with Haar Cascade Classifier and Histogram of Oriented Gradients

Abstract

With the emergence of Internet of Things (IoT) along with its development of advanced authentication, both security and remote monitoring have become imperative as well as essential, and the need for smarter security systems has only been growing. The traditional system needs an individual to use a key or an identification (ID) card or a password to access the security doors. However, they have many limitations such as keys can be forged, recreation of ID cards and passwords can be stolen. To overcome, the existing system issues, a novel approach is proposed with the design and development of face authenticated web-based smart door lock control system using facial recognition and remotely monitoring the door. In this proposed system OpenCVs self-trained Haar Cascade Classifier along with Histogram of Gradient is used for face Recognition. Door will be unlocked when users face is recognised else will remain closed. In case an unauthorised person is found, the time of intrusion and the intruders image will be captured and sent to a separate server on discord, so that the user or the admin can view them at their convenience. The main usage of this system is to assist users for improvement of the door security of sensitive locations by using face recognition and is also designed by considering the physically challenged persons also.