Person Verification Based on Multimodal Biometric Recognition

Nowadays, person recognition has received significant attention due to broad applications in the security system. However, most systems are implemented based on unimodal biometrics such as face or voice recognition. Biometric that adopted have limitations, mainly when data contains outliers and corrupted datasets. Multimodal biometric grab researchers’ consideration their superiority, better than system outstanding efficiency. Therefore, multimodal fingerprint is developed this paper. First, Convolutional Neural Network (CNN) ORB (Oriented FAST Rotated BRIEF) algorithm. Next, two features fused by using match score level fusion Weighted Sum-Rule. The verification process matched if greater pre-set threshold. algorithm extensively evaluated UCI Machine Learning Repository Database datasets, including one real dataset with state-of-the-art approaches. proposed method achieves a promising result