Capacity analysis of biometric hashing methods

We address the capacity analysis problem for the biometric hashing methods in this work. To the best of our knowledge, there is no work on this topic in the literature up to now. We develop an information theoretic capacity analysis method for biometric hashing methods. The proposed method depends on the within-class noise resilience which is analogues to the variations of the biometric data belonging to any user in the system. With the proposed method, we can also estimate the maximum number of users that a biometric hashing system can reliable accommodate. Besides, we test the performance of the proposed method with two different face image databases. Thus, we also experimentally estimate the maximum number of users that a biometric hashing method can handle. In order to compare the results with the another performance metric, we calculate the equal error rate of the biometric hashing methods as well.