A non-invertible Randomized Graph-based Hamming Embedding for generating cancelable ﬁngerprint template q Zhe Jin a,b , Meng-Hui Lim c , Andrew Beng Jin Teoh d,⇑ , Bok-Min Goi a a Faculty of Engineering and Science, Universiti Tunku Abdul Rahman, Kuala Lumpur, Malaysia b Department of Computer Engineering, Anhui Wonder University of Information Engineering, He Fei, China c Department of Computer Science, Hong Kong Baptist University, Kowloon Tong, Hong Kong d School of Electrical and Electronic Engineering, College of Engineering, Yonsei University, Seoul, South Korea article info Article history: Received 29 June 2013 Available online 26 February 2014 Keywords: Fingerprint template protection Non-invertible transform Randomized Graph-based Hamming Embedding Cancelable ﬁngerprint template abstract Biometric technology is likely to provide a new level of security to various applications. Yet if the stored biometric template is compromised, invasion of user privacy is inevitable. Since biometric is irreplaceable and irrevocable, such an invasion implies a permanent loss of identity. In this paper, a ﬁngerprint template protection technique is proposed to secure the ﬁngerprint minutiae. Remarkably, by incorporat- ing Randomized Graph-based Hamming Embedding (RGHE), the generated binary template can be strongly protected against inversion. The proposed method adopts a minutiae descriptor, dubbed as minutiae vicinity decomposition (MVD) to derive a set of randomized geometrical invariant features together with random projection. The discrimination of randomized MVD is then enhanced by User- speciﬁc Minutia Vicinities Collection scheme and embedded into a Hamming space by means of Graph-based Hamming Embedding. The resultant binary template enjoys four merits: (1) strong conceal- ment of the minutia vicinity, thus effectively protects the location and orientation of minutiae. (2) Well preservation of the discriminability of MVD in the Hamming space with respect to the Euclidean space without accuracy performance degradation. (3) Template is revocable due to user-speciﬁc random projection. (4) Speedy matching attributed to bit-wise operations. Promising experimental results on FVC2002 database vindicate the feasibility of the proposed technique. Ó 2014 Elsevier B.V. All rights reserved. 1. Introduction With the widespread deployment of biometric systems, using biometrics for identity veriﬁcation or identiﬁcation has raised much public concern about the security and privacy of biometric data over the last decade. Public worry about violation of user privacy is not uncommon, since the biometric data is inextricably bound to one’s identity and a compromise would lead to a perma- nent loss of their identity. For these inextricable mazes, a biometric system with strong template protection needs to be designed. In general, the design criteria for biometric template protection scheme are [1,2]:  Diversity. Cross-matching between templates from the same user across different applications must be prevented.  Cancelability. A new template can be reissued once the old tem- plate is compromised.  Non-invertibility. It should be computationally infeasible to derive the original biometric template from the protected tem- plate and the helper data.  Performance. The accuracy performance of an unprotected sys- tem should be preserved or improved. The template protection methods proposed in literature can be broadly divided into two categories, namely, feature transforma- tion approach (or cancelable biometrics) and biometric cryptosys- tem [3]. Biometric cryptosystem serves the purpose of either securing the cryptographic key using biometric feature (key binding) or directly generating the cryptographic key from biomet- ric feature (key generation) [3]. For key binding approach, two well- known instances, fuzzy commitment and fuzzy vault, are proposed by Juels and Wattenberg [4] and Juels and Sudan [5] respectively. On the other hand, Dodis et al. [6] introduces the key generation primitives, known as secure sketch and fuzzy extractor. http://dx.doi.org/10.1016/j.patrec.2014.02.011 0167-8655/Ó 2014 Elsevier B.V. All rights reserved. q This paper has been recommended for acceptance by Ajay Kumar. ⇑ Corresponding author. Tel.: +82 2 2123 5772; fax: +82 2 313 2879. E-mail address: bjteoh@yonsei.ac.kr (A.B.J. Teoh). Pattern Recognition Letters 42 (2014) 137–147 Contents lists available at ScienceDirect Pattern Recognition Letters journal homepage: www.elsevier.com/locate/patrec