Fingerprint Reference Point Detection Based on High Curvature Points
The problem considered in this paper is associated with the image processing and the design of algorithm which allows selecting reference point on the fingerprint image. The reference point is used to align between the fingerprints in the fingerprint auth
- PDF / 1,908,760 Bytes
- 10 Pages / 439.37 x 666.142 pts Page_size
- 74 Downloads / 205 Views
Abstract. The problem considered in this paper is associated with the image processing and the design of algorithm which allows selecting reference point on the fingerprint image. The reference point is used to align between the fingerprints in the fingerprint authentication systems faster than the conventional techniques. The reference point is the point with maximum curvature on the friction ridge, which is usually located in the central fingerprint area. Fingerprint homogeneous ridges are extracted from the image and then are processed by the IPAN99 algorithm which allows to detect curvatures of these lines. The experimental results on datasets of FVC2000, FVC2002, FVC2004 and NIST, show the high efficiency and satisfactory accuracy of the proposed algorithm. Proposed solution allows detecting reference points more precisely than other algorithms.
Keywords: Biometrics algorithm
1
·
Fingerprint
·
Reference point
·
IPAN
Introduction
Nowadays, people require much higher information security. The traditional identification methods, such as password, magnetic card, PIN, and even personal signature are already insufficient and far from being able to meet today’s environment. Biometrics is a personal identification technology based on an analysis of physiological or behavioral features of a given person [5,9,10]. In recent years, thanks to continuous technological development, a gradual increase of biometric techniques can be observed [2,14]. One of the most popular biometric techniques is fingerprint recognition. Fingerprint is a set of pattern of ridges, valleys (furrows) as well as characteristics that occur at minutiae points (ridge bifurcation or a ridge ending). These patterns are disclosed using inked impression on a paper or sensors (digital scanners). In police laboratories, to reveal difficult (latent) fingerprints magnetic powders, chemical baths and multi-impulse time resolved luminescence are also c Springer International Publishing Switzerland 2016 Y. Tan and Y. Shi (Eds.): DMBD 2016, LNCS 9714, pp. 538–547, 2016. DOI: 10.1007/978-3-319-40973-3 55
Fingerprint Reference Point Detection Based on High Curvature Points
539
applied. Fingerprints may come from different places such as crime scenes, police stations, biometric systems and many others. Fingerprint recognition technique is well known for many years and it is still developed and improved. Popularity of this method follows from their high efficiency, acceptability and applications in forensics. In practice, fingerprint analysis is a task from image processing domain because fingerprints traces are always converted to the digital images. Currently, fingerprint matching method can be classified into several categories [15]: minutiae based methods, where minutiae such as ridges endings and ridge bifurcations in two fingerprints are compare, ridge based methods where only ridges are matched, texture based methods in which the texture features of two fingerprints are matched, and neural network based methods. The state-of-the-art fingerprint minutiae based matchi
Data Loading...
