A Real-time Person Identity Detection System Using Machine Learning

Abstract

This research study presents an automated real-time background face recognition system for a large dataset of human faces. This is very difficult because background subtraction is still an issue in live images. Addition to this there are huge features in human face image in terms of eye, nose, head, lip, etc. The proposed system simplifies many of the facial recognition features. It utilizes AdaBoost with cascade to detect human faces in real-time. The matched face is then used to Identify a person. The real-time security and automation system is based on human face recognition, and it uses a simple and fast algorithm that achieves high accuracy. We have accuracy of 92% by using Adaboost Algorithm.