PROPOSE EUCLIDEAN DISTANCES CODECTION METHOD TO GUARANTEE USER PRIVACY IN FINGERPRINT RECOGNITION

Abstract

As technology develops, password authentication contains many security risks and is outdated. In the current authentication methods, the fingerprint authentication method is based on the identification of features, thereby comparing and identifying an individual. One of the methods for fingerprint recognition is using Euclidean distance. The problem is how when the user can provide a database that meets the algorithm used in the authentication model, giving accurate results without revealing information about the Euclidean distance when system reference. The author's solution uses the Paillier public cryptosystem system; encrypt the Euclidean distance for the authentication sample before entering the system for comparison; Experimental results encode the sample by Paillier coding, then check the correctness by calculating, comparing the results with the usual calculation. From there, the conclusion of the proposed algorithm is made.