Abstract
This paper presents a new framework for a biometric cryptosystem in which a cryptographic key is concealed with biometric modalities. In this paper, the candidate biometric modality is secured using two functions: 1) BCH encoding, which delivers the parity-code stored for the alignment of the query biometric template and 2) the Hash function to compute hash-code in order to safeguard its integrity. The cryptosystem is formed by creating two different cell-arrays. The hash-code is scattered on one cell-array by a randomly chosen column position, and the secret key is distributed over the second cell-array on the same position. The other cell-array locations are filled with the randomly generated chaff vectors. The parity-code is then jumbled up using a regenerative XORCoding in order to hide it from unauthorized access. At the unlocking stage, the parity-code is regenerated using XORcode and used to align the query template to the original one. If the hashed-code computed from the aligned template can locate the correct locations of the original hashed-code from the feature-array, it can get back the secret-key from the key-array and the correspondingly secret message. The proposed algorithm is implemented and evaluated in two modes: 1) unimodal and 2) multimodal. The experimental results from publicly available databases confirm the superiority of the multimodal cryptosystems over the unimodal cryptosystems.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have