Mobile Networks and Management. 9th International Conference, MONAMI 2017, Melbourne, Australia, December 13-15, 2017, Proceedings

Research Article

A Cancellable Ranking Based Hashing Method for Fingerprint Template Protection

Download
352 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-90775-8_30,
        author={Zhe Jin and Jung Hwang and Soohyung Kim and Sangrae Cho and Yen-Lung Lai and Andrew Teoh},
        title={A Cancellable Ranking Based Hashing Method for Fingerprint Template Protection},
        proceedings={Mobile Networks and Management. 9th International Conference, MONAMI 2017, Melbourne, Australia, December 13-15, 2017, Proceedings},
        proceedings_a={MONAMI},
        year={2018},
        month={5},
        keywords={Cancellable biometrics Fingerprint recognition Rank hashing},
        doi={10.1007/978-3-319-90775-8_30}
    }
    
  • Zhe Jin
    Jung Hwang
    Soohyung Kim
    Sangrae Cho
    Yen-Lung Lai
    Andrew Teoh
    Year: 2018
    A Cancellable Ranking Based Hashing Method for Fingerprint Template Protection
    MONAMI
    Springer
    DOI: 10.1007/978-3-319-90775-8_30
Zhe Jin1,*, Jung Hwang2,*, Soohyung Kim2,*, Sangrae Cho2,*, Yen-Lung Lai1,*, Andrew Teoh3,*
  • 1: Monash University Malaysia
  • 2: Electronics and Telecommunications Research Institute (ETRI)
  • 3: Yonsei University
*Contact email: jin.zhe@monash.edu, videmot@etri.re.kr, lifewsky@etri.re.kr, sangrae@etri.re.kr, lai.yenlung@monash.edu, bjteoh@yonsei.ac.kr

Abstract

Despite a variety of theoretical-sound techniques have been proposed for biometric template protection, there is rarely practical solution that guarantees non-invertibility, cancellability, non-linkability and performance simultaneously. In this paper, a cancellable ranking based hashing is proposed for fingerprint template protection. The proposed method transforms a real-valued feature vector into an index code such that the pairwise-order measure in the hashed codes are closely correlated with rank similarity measure. Such a ranking based hashing offers two major merits: (1) Resilient to noises/perturbations in numeric values; and (2) Highly nonlinear embedding based on the rank correlation statistics. The former takes care of the accuracy performance mitigating numeric noises/perturbations while the latter offers strong non-invertible transformation via nonlinear feature embedding from Euclidean to Rank space that leads to toughness in inversion yet still preserve accuracy performance. The experimental results demonstrate reasonable accuracy performance on benchmark FVC2002 and FVC2004 fingerprint databases. The analyses justify its resilience to inversion, brute force and preimage attack as well as satisfy the revocability and unlink ability criteria of cancellable biometrics.