Author Topic: A Novel Method for Fingerprint Core Point Detection  (Read 2968 times)

0 Members and 1 Guest are viewing this topic.

content.writer

  • Newbie
  • *
  • Posts: 48
  • Karma: +0/-0
    • View Profile
A Novel Method for Fingerprint Core Point Detection
« on: April 23, 2011, 01:37:21 pm »
Quote
Author : Navrit Kaur Johal, Prof. Amit Kamra
International Journal of Scientific & Engineering Research, IJSER - Volume 2, Issue 4, April-2011
ISSN 2229-5518
Download Full Paper - http://www.ijser.org/onlineResearchPaperViewer.aspx?A_Novel_Method_for_Fingerprint_Core_Point_Detection.pdf

Abstract- Fingerprint recognition is a method of biometric authentication that uses pattern recognition techniques  based on high-resolution fingerprints images of the individual. Fingerprints have been used in forensic as well as commercial applications for identification as well as verification. Singular point detection is the most important task of fingerprint image classification operation. Two types of singular points called core and delta points are claimed to be enough to classify the fingerprints. The classification can act as an important indexing mechanism for large fingerprint databases which can reduce the query time and the computational complexity. Usually fingerprint images have noisy background and the local orientation field also changes very rapidly in the singular point area. It is difficult to locate the singular point precisely. There already exists many singular point detection algorithms, Most of them can efficiently detect the core point when the image quality is fine, but when the image quality is poor, the efficiency of the algorithm degrades rapidly. In the present work, a new method of detection and localization of core points in a fingerprint image is proposed.

Index Terms—Core Point, Delta Point, Smoothening, Orientation Field, Fingerprint Classes.

Introduction
Fingerprints have been used as a method of identifying individuals due to the favorable characteristics such as “unchangeability” and “uniqueness” in an individual’s lifetime. In recent years, as the importance of information security is highly demanded, fingerprints are utilized for the applications related to user identification and authentication. Most Automatic Fingerprint Identification systems are based on local ridge features; ridge ending and ridge bifurcation, known as minutiae The first scientific study of the fingerprint was made by Galton who divided fingerprint into three major classes: arches, loops, and whorls. Henry, later refined Galton’s classification by increasing the number of classification. Henry’s classification is well-known and widely accepted. Henry’s classes consist of: arch, tent arch, left loop, right loop and whorl .

At a global level the fingerprint pattern exhibits the area that ridge lines assume distinctive shapes. Such an area or region with unique pattern of curvature, bifurcation, termination is known as a singular region and is classified into core point and delta point. The singular points can be viewed as the points where the orientation field is
discontinuous.
Core points are the points where the innermost ridge loops are at their steepest. Delta points are the points from which three patterns i.e. loop, delta and whorl deviate. Definitions may vary in deferent literatures, but this definition of singular point is the most popular one.  Figure 1 below represents the core and delta points.

 
Fig 1. The Core and Delta Points on a fingerprint image ( View Full Paper for Figure. )

This paper is organized as follows. In section 2, are discussed the different types of fingerprints. In Section 3 is explained the drawbacks with the existing techniques of core point detection. Section 4 focuses on the problem solution. In section 5, the core point is extracted using the proposed algorithm. The experimental results performed on a variety of fingerprint images are discussed in section 6 and the conclusion and future scope is discussed in Section 7.
2  FINGERPRINT CLASSES
The positions of cores and deltas are claimed to be enough to classify the fingerprints into six categories, which include arch, tented arch, left-loop, right-loop, whorl, and twin-loop.
•   Loops constitute between 60 and 70 per cent of the patterns encountered. In a loop pattern, one or more of the ridges enters on either side of the impression, recurves, touches or crosses the line of the glass running from the delta to the core, and terminates or tends to terminate on or in the direction of the side where the ridge or ridges entered. There is exactly one delta in a loop. Loops that have ridges that enter and leave from left side are called the Left Loops and loops that have ridges that enter and leave from right side are called the Right Loops. In twin loops the ridges containing the core points have their exits on different sides.

•   In a whorl, some of the ridges make a turn through at least one circuit. Any fingerprint pattern which contains 2 or more deltas will be a whorl pattern
•   In arch patterns, the ridges run from one side to the other of the pattern, making no backward turn. Arches come in two types, plain or tented. While the plain arch tends to flow rather easily through the pattern with no significant changes, the tented arch does make a significant change and does not have the same easy flow that the plain arch does.
In short, while classifying the fingerprints, we can make the assumption that if a pattern contains no delta then it is an arch, if it contains one (and only one) delta it will be a loop and if it contains 2 or more it will always be a whorl. If a pattern does contain more than 2 deltas it will always be an accidental whorl.

 
Fig 2. Classes of fingerprint (a)Arch, (b)Tented Arch, (c) Right Loop (d) Left Loop, (e) Whorl and (f) Double Loop (The double loop type is sometimes counted as whorl) 
Fingerprint friction ridge details are generally described in a hierarchical order at three levels, namely, Level 1 (pattern), Level 2 (minutiae points) and Level 3 (pores and ridge shape).Automated fingerprint identification systems (AFISs) employ only Level 1 and Level 2 features. No two fingerprints are alike, but the pattern of our fingerprint is inherited from close relatives and people in our immediate family. This is considered "level 1 detail." The detail of our actual finger and palm print is not inherited. This is considered "level 2 and 3 level detail" and is used to identify fingerprints from person to person.
The following figure briefly explains the three types of levels of details in our fingerprint:

   Fig 3.  Level 1 , Level 2 and Level 3 Details ( View Full Paper for Figure. )

In this paper, we propose a new core point detection method which can precisely localize the core point and does not require further post processing as well. The proposed method only concentrates on the core point detection as most of the ridge characteristics e.g ridge endings and ridge bifurcations are present in the core block (centre).

3  PROBLEM FORMULATION
The existing techniques used for detection of core point do not produce good results for noisy images. Moreover, they may sometimes detect spurious core point due to the inability to work efficiently for noisy images. Also techniques like Poincare Index fail for Arch type of Image. The aim of proposed algorithm is to formulate a more accurate core point determination algorithm which can produce better localization of core points avoiding any spurious points detected producing robust results for all types of fingerprints that have been discussed in this paper.

Read More: http://www.ijser.org/onlineResearchPaperViewer.aspx?A_Novel_Method_for_Fingerprint_Core_Point_Detection.pdf