Selecting distinctive features to improve performances of multidimensional fuzzy vault scheme
CCBR'11 Proceedings of the 6th Chinese conference on Biometric recognition
| Hi-index | 0.02 |
Fuzzy vault scheme is one of the most popular secret sharing mechanisms. In the scheme two people could share the secret concealed in the vault as long as two data sets A and B overlap with each other substantially. Because the two sets A and B are not required to be identical, many researchers consider the fuzzy vault scheme as one of biometric cryptosystems and try to implement it based on biometric feature data. However, the fuzzy vault scheme is not applicable to biometric feature data, because the scheme is designed based on set differences metric while in biometric systems Euclidean distance metric is often used to measure the similarity between two feature vectors. To improve the traditional fuzzy vault scheme to be applicable to biometric feature data, we propose a multidimensional fuzzy vault scheme, in which data matching could be performed the same as in biometric systems. A simple implementation of proposed multidimensional fuzzy vault scheme is given in this paper. Experimental results based on HA-BJTU palmprint database show the feasibility of proposed multidimensional fuzzy vault scheme.