Fingerprint Image Hashing Based on Minutiae Points and Shape Context

Abstract

Fingerprint minutiae is the unique representation of fingerprint image feature points as terminations and bifurcations. Therefore, generating a hash signature from these feature points will unarguably meet the desired properties of a robust hash signature and which will accurately fit in for fingerprint image content authentication purposes. This article proposes a novel minutiae and shape context-based fingerprint image hashing scheme. Fingerprint image minutiae points were extracted by incorporating their orientation and descriptors, then embedded into the shape context-based descriptors in order to generate a unique, compact, and robust hash signature. The robustness of the proposed scheme is determined by performing content preserving attacks, including noise addition, blurring and geometric distribution. Efficient results were achieved from the given attacks. Also, a series of evaluations on the performance comparison between the proposed and other state-of-art schemes has proven the approach to be robust and secure, by yielding a better result.