The two main types of template security schemes are the feature transformation approach and the biometric cryptosystems approach. In feature transformation, a transformation function is applied to an original model, and only the transformed model is stored in the database. The parameters of the transform function are usually derived from a password or random key. When the query model arrives, the same transformation function is applied to the model and it is transformed. Now the transformed model is checked for matching in the database. Depending on the characteristics of the transformation function, they are further divided into two types, namely salting and non-invertible transformations. Say no to plagiarism. Get a tailor-made essay on "Why Violent Video Games Shouldn't Be Banned"? Obtain an original essay In salting the transformation function is invertible, that is, if the adversary manages to get hold of the key and the transformed model then he can obtain the original model. Therefore the security of the salting process depends on the protection of the key or password used in the transformation, while in the case of non-invertible transformations the transformation function is a one-way process and it is computationally very difficult to recover the original model value from the transformed model even if the key is sacrificed. Initially, biometric cryptosystems were created to protect a cryptographic key using biometric traits or by creating cryptographic keys from biometric templates. Furthermore, the same techniques were used to protect biometric templates as well. In biometric cryptosystems, some public information about the model, called supporting data, is usually stored. Therefore biometric cryptosystems are also called supporting data-based methods. The supporting data cannot be used to revoke the original model or does not reveal significant information about the original model but is used during the matching process to extract the cryptographic key from the query model. In this case, matching is an indirect process that is performed by verifying the correctness of the extracted cryptographic key. Biometric cryptosystems are further classified into key binding and key generation approaches which depend on how the supporting data is obtained. When the helper data is obtained by associating a key (which is independent of the biometric model) to the original biometric model, then it is called a Key Binding biometric cryptosystem. It should be noted that given only supporting data it is very difficult to obtain the original model. In a key binding system, matching is done by retrieving the key from the supporting data along with the help of the biometric model. On the other hand, if the supporting data is obtained only from the biometric model and the cryptographic key is also derived from the supporting data and also from the query biometric model, then it is called Biometric Key Generation Cryptosystem. Both biometric cryptosystems and pattern transformation systems have their advantages and disadvantages. Templates under transformation can be easily revoked by changing the password or key. Sophisticated matching algorithms and matching devices can be designed that can robustly handle intra-user variations in the transformer biometric model, due to fewer restrictions placed on the matching algorithms that can be used in the transform domain. This also reduces the error rates of the biometric system. However, it is difficult to measure.