Biometric Access Using Image Processing Semantics

Abstract

We propose a model-based approach for accessing a magnetic lock in the door using face recognition. We Implement Face Recognition by using image processing semantics. The Facial recognition uses facial landmarks such as chin, eyebrow, nose, eyes and lip to encode the user’s face. The encoded signal is sent to the lock via WIFI. We interfaced the facial recognition module through an android app and implemented the backend using python server. Our approach here is by using KNN algorithm which is one among the various machine learning algorithm. The K-nearest neighbor algorithm is used to classify the facial landmarks which is calculated by Euclidian distance formula which calculates the distance between the markers. The Face recognition module runs through python which is used for connectivity between the magnetic lock and backend. The magnetic lock is implemented using Node MC, switch, electromagnet and a power supply.