Reliable Identity Management System Using Raspberry Pi

摘要

Biometric authentication, the process of verification using unique biological characteristics, is considered one of the most accurate technologies for authentication. It compares two or more biometric data to identify the differences among them properly. Based on the comparison between individual’s biometric features and stored authorized features, it makes the decision of matched or mismatched. If the given sample of the biometric data matches with the stored one, authorization is confirmed. Nowadays, many real-time applications use biometric authentication to meet security purposes fully. However, it is difficult to implement fingerprint recognition system specially on Linux based computer. A well-known, less costly embedded computer called Raspberry Pi is used to build the entire system. Fingerprint enrollment, fingerprint recognition, photo capture, QR code generation as well as QR code reading were entirely done using python and python associated libraries. For storage management, we have used MySQL as it is convenient and open-source. This paper discusses about a quality authentication model which is highly capable and cost- effective. The authentication process consists of extracting the biometric features with ORB algorithm, matching those features using FLANN matcher, generating and validating QR code based on encrypted personal information accurately. Here, AES algorithm is used for encrypting personal data which is then converted and stored in form of QR code. Final authentication process involves both fingerprint and QR code validation. Performing two factor authentication affixes more reliability to the system.