As stated in ISO/IEC JTC1 SC37, biometrics is defined as the automated recognition of individuals based on their behavior; and biological characteristics. As discussed in the ------- (Introduction to Biometrics) section, the main processes involved in biometric recognition are registration and verification or identification in general called recognition. During the registration process, the characteristics of the acquired biometric entity are converted into a template, to be stored in the database for further correspondence. In the recognition phase, the patterns obtained from the real-time biometric entity features are compared with the stored pattern. The result of the matching process involved in recognition is accept or reject[1]. It has long been believed that biometric signals or data cannot be reconstructed from stored templates, but Cappelli et.al and Ross et.al[2] [3] have shown that believing that reconstructing biometric data from templates is possible . Several standard encryption algorithms have been used to protect biometric data. However, this attempt leaves the biometric templates exposed during each authentication attempt[4]. Homomorphic and asymmetric encryption schemes [5] [6] [7] are also some exceptions. Conventional cryptographic systems are based on the possession of secret keys, and key management is performed using second-level authentication entities such as password[8]. The decryption keys in this scheme could be obtained using password. To overcome the drawbacks of existing schemes, biometric template protection schemes commonly called biometric cryptosystems, also called Helper data-based schemes, are proposed. Two main requirements for the protection of biometric information according to the ISO/IEC FCD 24745 standard are: • Irreversibility The reconstruction of the original biometric template from the stored template should be made difficult while the construction of the protected biometric template from the stored template should be made easy. • Non-linkability Several different versions of protected biometric templates can be generated based on the same biometric data, a phenomenon called renewability, while the protected templates do not have to match, a phenomenon called diversity. Biometric cryptosystems hereinafter referred to as BCS are designed to securely bind a digital key to a biometric template or generate a digital key [9]. BCS paves the way for the growth of biometric key issuance and biometric template protection [10] [11]. BCS is more difficult to spoof, copy, share and distribute biometric data than passwords [1]. Conventional biometric cryptosystems perform fuzzy comparisons by applying decision thresholds. The decision thresholds are obtained based on the distribution of scores between real subjects and impostor subjects, while BCS provides only stable keys as output, necessary for the correspondence during the authentication phase.