An Efficient Fingerprint Image Encryption Algorithm

In this paper, we propose a fingerprint image encryption algorithm based on permutation and diffusion operations in wavelet domain in order to enhance the protection of fingerprint-based systems against replay attacks. These attacks have been identified and located in different points by Ratha et al. First, one-level Lifting Wavelet Transform Integerto- Integer is performed to the original fingerprint image. The approximation and details sub-bands are then partitioned into blocks and permuted using a permutation key. After that, for each sub-band the Rubik’s cube principle presented in (Loukhaoukha et al., 2012) is applied. Then, the values of rows and columns of the modified sub-bands are changed using XOR operator. Finally, the encrypted image is constructed by ordering the encrypted sub-bands. Experimental tests and security analysis have been carried out on three fingerprint images, taken from Fingerprint Verification Competition, FVC 2000 database. The obtained results clearly show the robustness of the proposed encryption algorithm against common attacks, namely exhaustive, differential and statistical attacks and also reveal the high security level achieved by theproposed algorithm.

Keywords:

Image Encryption , Block Permutation , Wavelet Transform , Fingerprint Image