Cylindrical Curve for Contactless Fingerprint Template Securisation

1. Introduction

Biometrics identifies human characteristics and traits. The fingerprint remains the most used characteristic during human authentication because it works well and is unique. The minutiae are the most adopted representation. Although minutiae are the most adopted representation, it is exposed to several dangers that several researches have proven. The architecture of the figure 2 gives several types of attacks. The literature reviews vulnerabilities and attacks against a biometric system are described in several ways and based on the opinion of several authors in (Ratha et al, 2001; Ratha et al, 2003;Cukik et al, 2005; Adler et al, 2005; Jain et al, 2006; Roberts et al, 2008; Jain et al, 2008) . Eight levels of attacks have been identified, namely:

Figure 1.

Secured contactless cylindrical fingerprint template generation

Figure 2.

Fish bone model to categorize vulnerabilities of fingerprints template

• ü Presentation attacks: fingerprints are presented at the entrances after having reproduced them;

• ü hacking and use of fingerprint data after bypassing the sensor;

• ü Usurped features are substituted for the original;

• ü Tampering with the correspondence module to use false functionality;

• ü Data replay attacks;

• ü Replacement of the characteristics module by a trojan horse;

• ü Spying on the channel between the feature extraction module and the classifier by an opponent in order to record the original model and replay it.

Each attack is dependent on a set of resources and constraints which makes it possible to classify it according to its acuity or its severity. The attack on the fingerprint templates stored in the comparison module, should be considered a major threat to authentication systems and one of the most formidable attacks against fingerprint-based authentication systems digital.