Entry restricted security system for a smart home

Abstract

This report represents a new personal identification method for entry restricted security system. In this proposed method skeleton recognizing feature of Kinect is being used to identify human. Human identification has been done by comparing 19 body segment’s length and 8 angles of different body joints with the saved lengths and angles. Fuzzy C–means (FCM) algorithm is used to compare new data with the data in the database. The system will decide whether to restrict or allow any person depending on the comparison result of FCM algorithm. Experimental results show that the proposed method exhibited 100% accuracy for our tested dataset.