Fingerprint Identification Scheme Based on Distribution Density

Abstract. Traditional fingerprint identification is adopting minutiae point as a template， but this exist template leaked danger. Based on the distribution density of minutiae point， this paper deeply researches on how to use the distribution density of minutiae point as the template of fingerprints， avoiding directly storing minutiae point data， and ensuring the safety of fingerprint template. At the same time， we proposed a fingerprint matching algorithm based on this template. The experimental results show that the matching algorithm is an effective identification scheme.

Key words： Fingerprint matching， Minutiae point matching， Genetic algorithm

Introduction

The fingerprint is human pattern which formed by concave and convex skin texture. With its characteristics of uniqueness， the fingerprint is widely used in identification， such as forensic probe， personal identity， etc.

In the early， people used global structure features to match fingerprint. At present， the matching algorithm is mainly based on minutiae point matching. Minutiae point are some singular point in the fingerprint image， it mainly includes the ridge line endpoint， ridges bifurcation points， isolation， holes， and etc. Some of these points are only a small part， such as bridge， isolated point， and they are difficult to be detected or easily regarded as a noise during fingerprint image processing. On the contrary， the endpoint and bifurcation point account for 82.6% of the total number of minutiae point. So we choose the ridge line endpoint and ridges bifurcation point as the matching minutiae point.

There are many of algorithms based on minutiae point matching， such as point pattern matching， the structure of the minutiae point matching， the matching of boundary box. Yuan-Yuan Zhang and Xiao-Jun Jing proposed an algorithm based on boundary box with genetic algorithm [2]； Yuan Gao and Ming-Zhi Zhang proposed an algorithm based on Similar Vector Triangle [3]. Although these algorithms can provide a better matching result， there is one of the same faults， which all need to store the minutiae point of the fingerprint as a template. With the application of fingerprint identification is more and more popular， the fingerprint template leakage problem is increasingly serious， and the leak is unrecoverable. So， the researchers made a lot of research in protecting the fingerprint template， including the theory of fuzzy vault， fuzzy extract， etc. Those solutions mainly concentrated on the theoretical study， but it lack relevant implementation， so the actual project safety remains to be proven.

We propose a security fingerprint identification scheme in this paper. Because there isn’t save distribution information of minutiae point， attacker can’t steal fingerprint characteristics of minutia data directly， so we can effectively solve the traditional fingerprint template leak problem. On this basis， the matching algorithm based on genetic algorithm is proposed. In this paper， we use the new matching algorithm conduct an experiment， and error receiving rate and false rejection rate results is in an acceptable range， so we think this is a kind of effective matching algorithm.

This paper is organized as follows. In Section II， we introduce the distribution density of minutiae point and illustrate the different distribution density between different fingerprints can be distinguished. On this basis of this new fingerprint template， the matching algorithm based on genetic algorithm is proposed in section III. In section IV， the results of experiment are analyzed and summarized. Finally， conclusion remarks and future work are shown in Section V.

Distribution Density Of Minutiae And Distinguishable

It is well known that there are no two identical fingerprints in the world. Early before the computer， scholars proved the uniqueness of the fingerprints. In the modern fingerprint identification system， the minutiae point of the fingerprint is often used. On the model of distribution， this paper put forward a distribution density model based on minutiae point. The distribution density of minutiae point is a measure， which describes distribution density of the minutia. Larger the density value it is， the distribution of the minutiae point is more intense.

References

[1] Federal Bureau of Investigation， "The Science of Fingerprints： Classification and Uses". Washington， D.C.： U.S. Government Printing Office， 1984.

[2] Yuan-yuan Zhang and Xiao-jun Jing. “Fingerprint Matching Based on Fast Genetic Algorithm”. Computer Engineering， 2011， 37（24）.

[3] Jian-De Zheng， Yuan Gao， and Ming-Zhi Zhang， “Fingerprint Matching Algorithm Based on Similar Vector Triangle”， IEEE Trans. Inf. Forensics Security， vol. 2， no. 4， pp. 744–757， Dec. 2007.

[4] Bo Hong， Gang Rong and Tao Huang. “Fingerprint matching based on the genetic algorithm”. Tsinghua Science and Technology. 2001， 41（3）.

[5] X.Tan and B.Bhanu， “Fingerprint matching by genetic algorithms，” Pattern Recognit， vol. 39， no. 3， pp. 465-477，2006.

[6] Y.He，J.Tian，X.Luo，and T.Zhang， “Image enhancement and minutiae matching in fingerprint verification，” Pattern Recognit.Lett， vol.24，pp.1349-1360，2003.