Development of a New Cryptographic Construct Using Palmprint-Based Fuzzy Vault

Amioy Kumar, Ajay KumarEURASIP J. Adv. Signal Process 2009 · DOI

Abstract

The combination of cryptology and biometrics has emerged as a promising component of information security. Despite the current popularity of palmprint biometrics, there has not been any attempt to investigate its usage for the fuzzy vault. This paper investigates the possible usage of palmprint in a fuzzy vault to develop a user-friendly and reliable cryptosystem. We suggest the use of both symmetric and asymmetric approaches for encryption. The ciphertext of any document is generated by a symmetric cryptosystem; the symmetric key is then encrypted by an asymmetric approach. Further, Reed-Solomon codes are used on the generated asymmetric key to provide error tolerance during decryption. Experimental results suggest the viability of an automated palmprint-based key generation system.

Introduction

Traditional cryptographic systems rely on long, complex keys that are difficult for humans to memorize. Conversely, biometric security offers convenience but faces the "biometric dilemma": unlike a password, a biometric template cannot be changed if compromised. The "fuzzy vault" construct was proposed to bridge this gap, allowing a secret to be "locked" with noisy biometric data. While fingerprints have been the primary focus of fuzzy vault research, they often suffer from extraction difficulties in certain demographics and carry a social stigma. This research explores palmprints as a more feature-rich, stable, and user-friendly alternative for biometric key protection.

Proposed Hybrid Architecture

The core of the proposed system is a "double encryption" strategy designed to balance speed and security.

  1. Bulk Data Security: Large documents are encrypted using the Advanced Encryption Standard (AES), a high-speed symmetric algorithm.
  2. Key Management: The AES symmetric key is itself encrypted using RSA, an asymmetric algorithm.
  3. Biometric Locking: The critical RSA private key is not stored as a file or protected by a password; instead, it is embedded within a fuzzy vault constructed from the user's palmprint features.
The double encryption framework combining AES symmetric encryption with RSA asymmetric key management.

The Palmprint-Based Fuzzy Vault

The fuzzy vault serves as a secure container for the RSA private key. To handle the inherent variability in biometric scans (due to pressure or alignment), the system incorporates several technical layers:

  • Feature Extraction: The system applies a Discrete Cosine Transform (DCT) to 144 overlapping blocks of the palmprint image. This process generates a robust feature vector that captures the unique textures of the palm.
  • Error Correction: Reed-Solomon (RS) codes are applied to the secret key. This mathematical framework allows the system to reconstruct the original key even if the query palmprint differs slightly from the enrolled template.
  • Security via Chaff Points: To prevent brute-force attacks, the vault is populated with "chaff points"—random data that is indistinguishable from real biometric features but does not contribute to the key reconstruction.
The process of generating a fuzzy vault by combining Reed-Solomon coded keys with normalized palmprint features and random chaff points.

Experimental Evaluation

The system was tested using a database of 170 palmprint images from 85 users, captured in a realistic "peg-free" setup. Performance was measured using the Equal Error Rate (EER), where the probability of false matches (FAR) equals the probability of false rejections (FRR).

Key findings include:

  • An EER of 0.375% was achieved for a 307-bit RSA key length.
  • In specific configurations, the system reached a False Rejection Rate (FRR) of 0% with a minimal False Acceptance Rate (FAR) of 0.35%.
  • The results indicate that palmprint features provide higher "intraclass" stability and more unique data points than fingerprints, making them ideal for high-security fuzzy vaults.

Conclusion and Future Directions

This research demonstrates that palmprints are a viable and highly effective biometric for securing cryptographic keys. By decoupling the biometric from the key itself—using the palmprint only to "lock" a standard RSA key—the system provides revocability; if a key is compromised, a new one can be generated and re-locked without needing to change the user's biometric data.

Future work will focus on scaling the dataset to thousands of users to ensure "bank-grade" reliability and optimizing the computational overhead of generating large vaults with high densities of chaff points. Additionally, exploring multi-modal biometrics (e.g., combining palmprints with hand geometry) could further harden the system against sophisticated spoofing attempts.

Summary updated

BibTeX
@article{kumar2009development,
  author = {Amioy Kumar and Ajay Kumar},
  title = {{Development of a New Cryptographic Construct Using Palmprint-Based Fuzzy Vault}},
  journal = {EURASIP J. Adv. Signal Process.},
  year = {2009},
  doi = {10.1155/2009/967046},
}
Figure