Authentication of 3D fuzzy vault using invariant facial features

Abstract

Biometrics and Cryptography play independent vital roles in the field of security. Blend of these two technologies, produce possible high level security system viz., Biometric-Crypto system in which bio templates assist encryption and decryption process. The major issues associated with current cryptographic algorithms using key which are long and highly random are: a) the randomness and uniqueness provided by current mathematical algorithms is felt inadequate to the current commercial applications; b) most of the authentication mechanisms use passwords to release the correct decrypting key, but these mechanisms are unable to provide non-repudiation; c) most of the biometric authentication systems store multiple templates per user to account for variations in biometric data. Such systems suffer from storage space and computation overheads; d) protecting the biometric template stored in a database or a smart card; e) maintenance of absolute secrecy of the key. Traditional cryptographic commitment needs exact data that is, unique decryption key, where as the biometric keys are prone to random noises. Even if the knowledge from biometric systems is exact, transmission channels may introduce noise which leads to fuzziness. Further fuzziness may come from the variable nature of biometric data. Though same biometric template is analyzed during different acquisitions; the extracted biometric data will vary and cause fuzziness. Hence cryptographic systems must be made to accept some fuzziness regarding witness that is, it must be decrypted by a key that is nearer to encryption key. A fuzzy extractor addresses both error tolerance and non uniformity. It extracts a uniformly random string from a biometric input. The slight changes in the input will be tolerated and this does not affect the extracted string. The present work is aimed at addressing the above problems, and to suggest efficient solutions. The work mainly focuses on creation of fuzzy vault using the fuzzy facial features.