A Hybrid Biometric and Password-Based Dual Security System Implemented through MATLAB Simulation

Abstract

In an era marked by growing concerns over security breaches and identity theft, single-layer authentication systems such as numeric passwords or biometrics alone are no longer sufficient to ensure reliable protection. This study presents the design and simulation of a hybrid dual-level security system that integrates biometric face recognition with password-based verification to achieve enhanced access control. The first level of authentication employs the Principal Component Analysis (PCA) algorithm, implemented in MATLAB, to extract Eigenfaces from a training dataset and compare them against test images. This technique enables dimensionality reduction while preserving essential facial features, ensuring efficient and accurate recognition under controlled conditions. Once a face is successfully identified, the system advances to the second level, where a password or personal identification number (PIN) is verified through an 8051 microcontroller using the Edsim51di simulator. A correct match results in the activation of a motor, simulating door unlocking, while mismatches deny access. Experimental results demonstrate that the PCA-based module achieved a recognition accuracy of approximately 75.83%, while the password verification layer maintained near-perfect reliability. Together, these two mechanisms provide a defense-in-depth model that minimizes false acceptance, reduces unauthorized access, and enhances overall system robustness. The proposed hybrid framework offers a cost-effective and practical solution suitable for high-security applications such as ATMs, airports, and smart homes.