An enhanced fuzzy commitment scheme in biometric template protection
Biometric template protection consists of two approaches; Feature Transformation (FT) and Biometric Cryptography (BC). This research focuses on Key-Binding Technique based on Fuzzy Commitment Scheme (FCS) under BC approach. In FCS, the helper data should not disclose any information about the biomet...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | eng eng |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://etd.uum.edu.my/10637/1/Depositpermission_s901634.pdf https://etd.uum.edu.my/10637/2/s901634_01.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Biometric template protection consists of two approaches; Feature Transformation (FT) and Biometric Cryptography (BC). This research focuses on Key-Binding Technique based on Fuzzy Commitment Scheme (FCS) under BC approach. In FCS, the helper data should not disclose any information about the biometric data. However, literatures showed that it had dependency issue in its helper data which jeopardize security and privacy. Moreover, this also increases the probability of privacy leakage which lead to attacks such as brute-force and cross-matching attack. Thus, the aim of this research is to reduce the dependency of helper data that can caused privacy leakage. Three objectives have been set such as (1) to identify the factors that cause dependency on biometric features (2) to enhance FCS by proposing an approach that reduces this dependency, and (3) to evaluate the proposed approach based on parameters such as security, privacy, and biometric performance. This research involved four phases. Phase one, involved research review and analysis, followed by designing conceptual model and algorithm development in phase two and three respectively. Phase four, involved with the evaluation of the proposed approach. The security and privacy analysis shows that with the additional hash function, it is difficult for adversary to perform brute‐force attack on information stored in database. Furthermore, the proposed approach has enhanced the aspect of unlinkability and prevents cross-matching attack. The proposed approach has achieved high accuracy of 95.31% with Equal Error Rate (EER) of 1.54% which performs slightly better by 1.42% compared to the existing approach. This research has contributed towards the key-binding technique of biometric fingerprint template protection, based on FCS. In particular, this research was designed to create a secret binary feature that can be used in other state-of-the-art cryptographic systems by using an appropriate error-correcting approach that meets security standards. |
|---|