Evaluating Performance of SIFT Algorithm in Detection of the Hidden Fingerprint in E-Commerce Transactions

E-commerce is outcome of the development and globalization of information technology in all areas of human life. Data security and privacy play an important role in e-commerce. Biometric technology as a solution can be used for authentication in e-commerce transactions. But the theft of biometric information such as fingerprint is a challenging problem. Sending a confidential data with the fingerprint image can avoid this situation. In this paper, the RSA algorithm and the Diffie-Helman algorithm are used to encrypt the confidential data. Then the encrypted date is hidden in a fingerprint image. SIFT algorithm is applied to extract corresponding points in the two data sets from the original images and Stego image (already information embedded image) and performance of this descriptor in fingerprint matching is investigated. Results showed the SIFTalgorithm matched 88.23% and 96.15%, by applying the RSA algorithm and Diffie-Helman algorithm, respectively. Therefore, the steganography of a confidential data in the fingerprint image provides access of the authorized people to biometric information and the quality of fingerprint image does not change.