Implementation of Integrated Security System by Using Biometric Function in ATM Machine

摘要

The paper proposes an efficient approach for authentication/identffication of a valid person at ATM machine using some body parameters along with normal procedures which are currently used. Body parameters are taken through hidden camera installed in ATM and the extracted data is processed based on matching percentage with the stored data to identify/authenticate the valid person. If a person is found suspicious then appropriate action is triggered. For experimental work, two body parameters, i.e., fingerprint and face recognition have been taken to do the analysis for better result. Two databases, i.e., Olivetti Research Laboratory (ORL) face database and Extended Cohn-Kanade database (CK+) have been used in face recognition. For fingerprint analysis, stored data and test data have been compared to get the valid result. All experiments have been done in MATLAB and the result obtained by the proposed work is quite impressive.