Enhanced Thinning Based Finger Print Recognition

Published Date: 02 Nov 2017

Abstract— This paper is the implementation of fingerprint

recognition system in which the matching is done using the

Minutiae points. The methodology is the extraction &

applying matching procedure on the Minutiae points

between the sample fingerprint & fingerprint under

question. The main functional blocks of this system follows

steps of image thinning, image segmentation, Minutiae

(feature) point extraction, & Minutiae point matching.

The procedure of image thinning for the purpose of

decreasing the size of the memory space used by the

fingerprint image database.

Keywords— Minutiae; Fingerprint; Segmentation; Image

Thinning

I. INTRODUCTION

A finger print is the pattern of ridges & valleys; each

individual has unique fingerprints. The uniqueness of a

fingerprint is exclusively determined by the local ridge

characteristics & their relationship. The two most prominent

local ridge characteristics, called Minutiae, are the ridge

ending & the ridge bifurcation [1]. A good quality fingerprint

typically contains 40-100 minutiae [2, 3], as shown in the

figure 1.

Fig 1. Finger Print

Fortunately, controlled, scientific testing initiatives are not

limited within the biometrics community to fingerprint

recognition. Other biometric modalities have been the target

of excellent evaluation efforts as well. The (US) National

Institute of Standards and Technology (NIST) has sponsored scientifically-controlled tests of text-independent speaker recognition algorithms for a number of years and, more recently, of facial recognition technologies as well.

The goal of feature extraction in pattern recognition system

(in general) is to extract information from the input data that is

useful for determining its category. In the case of fingerprints

a natural choice are features based directly on the fingerprint

ridges and ridge-valley structure. However, the effectiveness

of a feature extraction depends greatly on the quality of the

images. Consequently, fingerprint image enhancement has

become a necessary and common step after image acquisition

and before feature extraction in most AFIS. Following,

binarization, feature extraction and matching algorithms are

executed on the enhanced image. The fingerprint enhancement

can be employed on, both the gray-scale, and binary images,

in spatial or frequency domain. In this paper we propose

method based on Enhanced thinning of Images of original

input gray-scale fingerprint image in frequency domain.

The rest of this paper is organized as follows. In Section II Background and Literature survey presented and its characteristics in frequency domain are briefly described. In Section III proposed methodology is explained. Parameters of proposed technique as well as results of enhancement obtained from available database sets are shown in Section IV, and a conclusion is given in Section V.

II. BACKGROUND AND LITERATURE SURVEY

Bana. S, with her colleague in 2011 presents a technique

which is based on Minutiae based matching. This approach

mainly depends on extraction of minutiae point from the

sample finger print images and then perform matching based

on the number of minutiae pairing among two fingerprints [4].

Andelija M, et.al in 2009 propose an adaptive filtering in

frequency domain in order to enhance fingerprint image. Due

to development of fast algorithms and power of modern

computer systems, the filtering is often done in frequency

domain. They propose two filter realizations for adaptive

filtering in frequency domain, where both of them enhance

fingerprint ridge-valley structure and attenuate existing noise

[5].

K.Thaiyalnayaki, with his colleague in 2010 propose an effective combination of features for multi-scale and multi-directional recognition of fingerprints. The features include standard deviation, kurtosis, and skewness. They apply the method by analyzing the finger prints with discrete wavelet transform (DWT) [6].

A. Jagna, et.al in 2010 proposes new parallel thinning

algorithm for binary images [7]. The proposed algorithm improves the ZS and LW algorithm’s problem that the excessive erosion and discontinuity in the thinned images. The algorithm executes thinning of input image through two iterations. presented a two-pass parallel binary image thinning algorithm that make the image one pixel thick and preserves the end points. This algorithm also ensures the 8-neighbour connectivity. The ZS and LW

algorithm can reduce the end points [8-9]. However, the

proposed algorithm shows the better performance and

produces more quality images than the previous algorithms.

Eun-Kyung Yun, Jin-Hyuk Hong and Sung-Bae Cho in

2005 propose an adaptive preprocessing method, which

extracts five features from the fingerprint images, analyzes

image quality with Ward’s clustering algorithm, and

enhances the images according to their characteristics. Fig.

2 shows the overview of the proposed system in this paper.

For fingerprint image quality analysis, it extracts several

features in fingerprint images using orientation fields, at

first. Clustering algorithm groups fingerprint images with the

features, and the images in each cluster are analyzed and

preprocessed adaptively [10].

Fig 2. Adaptive Enhancing of Finger Print Images

Madhuri and Richa Mishra in 2012 proposes a fingerprint

recognition technique which uses local robust features for

fingerprint representation and matching. The technique

performs well in presence of rotation and able to carry

out recognition in presence of partial fingerprints [11].

Fig 3. Finger Print based on Local robust features

Finger Print Matching Techniques:

The large number of approaches to fingerprint matching can be coarsely classified into three families.

• Correlation-based matching: Two fingerprint

images are superimposed and the correlation between

corresponding pixels is computed for different

alignments (e.g. various displacements and rotations).

• Minutiae-based matching: This is the most popular and

widely used technique, being the basis of the fingerprint

comparison made by fingerprint examiners. Minutiae

are extracted from the two fingerprints and stored as sets of

points in the two- dimensional plane. Minutiae-based

matching essentially consists of finding the

alignment between the template and the input minutiae sets

that results in the maximum number of minutiae pairings.

• Pattern-based (or image-based) matching:

Pattern based algorithms compare the basic fingerprint

patterns (arch, whorl, and loop) between a

previously stored template and a candidate fingerprint.

This requires that the images be aligned in the same

orientation. To do this, the algorithm finds a central point

in the fingerprint image and centers on that. In a pattern-

based algorithm, the template contains the type, size, and

orientation of patterns within the aligned fingerprint image.

The candidate fingerprint image is graphically compared

with the template to determine the degree to which they

match.

Issues with Existing techniques:

Most of the existing fingerprint techniques in literature are based on minutiae points which are represented using their co-ordinate locations in the image. When test fingerprint image is rotated with respect to enrolled image or partially available, these techniques face problem in matching due to change in the co-ordinate locations of the minutiae points and perform very poorly. These two cases are discussed below.

A. Rotated Fingerprint Matching:

An example of a rotated fingerprint image is shown in Figure 1(b). We can see that it is difficult to match

minutiae of two images because due to rotation, coordinate locations of all the minutiae points in Figure 1(b) with respect to Figure 1(a) are changed.

B. Partial Fingerprint Matching:

An example of partial fingerprint is given in Figure 2(b).

We can see that it is difficult to match minutiae of two

images because due to missing part of the fingerprint

coordinate locations of all the minutiae points in Figure

2(b) with respect to Figure 2(a) are changed.

fact ridges, still vary in intensity. So, binarization

transforms the image from a 256-level image to a 2-level

image that gives the same information. Typically, an

object pixel is given a value of "1" while a background

pixel is given a value of "0." Finally, a binary image is

created by coloring each pixel white or black, depending

on a pixel's label (black for 0, white for 1).

III. PROPOSED METHODOLOGY

The Fig 4.4 shows the proposed methodology. Basic steps

involves are written below:

1. Image Enhancement: The first step in the minutiae

extraction stage is Fingerprint Image enhancement.

Fingerprint Image enhancement is used to make the image

clearer for easy further operations. Since the fingerprint

images acquired from scanner or any other media are

not assured with perfect quality, those

enhancement methods, for increasing the contrast

between ridges and valleys and for connecting the false

broken points of ridges due to insufficient amount of ink,

are very useful for keep a higher accuracy to fingerprint

recognition. Originally, the enhancement step was

supposed to be done using the canny edge detector. But

after trial, it turns out that the result of an edge detector is an image with the borders of the ridges highlighted. For image enhancement we use:

a. Histogram Equalization

b. Fast Fourier Transform

2. Image Binarization: The binarization step is basically

stating the obvious, which is that the true information that

could be extracted from a print is simply binary;

ridges vs. valleys. But it is a really important step in the

process of ridge extracting, since the prints are taken as

grayscale images, so ridges, knowing that they’re in

3. Image Segmentation: In general, only a Region of

Interest (ROI) is useful to be recognized for each

fingerprint image. The image area without effective

ridges is first discarded since it only holds background

information and probably noise. Then the bound of the

remaining effective area is sketched out since the

minutiae in the bound region are confusing with those

false minutiae that are generated when the ridges are out of

the sensor. To extract the ROI, a two-step method is

used. The first step is block direction estimation and

direction variety check, while the second is done using

some Morphological methods. Used techniques are:

a) Block Direction Estimation

b) ROI extraction by Morphological Operation

4. Final Minutiae Extraction: Ridge Thinning is to

eliminate the redundant pixels of ridges till the ridges are

just one pixel wide. An iterative, parallel thinning

algorithm is used. In each scan of the full fingerprint

image, the algorithm marks down redundant pixels in each

small image window (3x3) and finally removes all those

marked pixels after several scans. The thinned ridge

map is then filtered by other Morphological operations

to remove some H breaks, isolated points and spikes.

In this step, any single points, whether they are single-

point ridges or single-point breaks in a ridge are eliminated and considered processing noise.

Minutia Marking: After the fingerprint ridge thinning,

marking minutia points is relatively easy. The concept of

Crossing Number (CN) is widely used for extracting the

minutiae.

In general, for each 3x3 window, if the central pixel is 1 and has exactly 3 one-value neighbors, then the central pixel is a ridge branch [Figure 4.1].

Fig 4.1

If the central pixel is 1 and has only 1 one-value neighbor, then the central pixel is a ridge ending [Figure4.2], i.e., for a pixel P, if Cn(P) = = 1 it’s a ridge end and if Cn(P) = = 3 it’s a ridge bifurcation point.

Image

Image

Segmentation

Fig 4.2

Fig 4.3 illustrates a special case that a genuine branch is

triple counted. Suppose both the uppermost pixel with value

1 and the rightmost pixel with value 1 have another neighbor outside the 3x3 window, so the two pixels will be marked as branches too, but actually only one branch is located in the small region. So a check routine requiring that none of the neighbors of a branch are branches is added.

Also the average inter-ridge width D is estimated at this

stage. The average inter-ridge width refers to the average

distance between two neighboring ridges. The way to

approximate the D value is simple. Scan a row of the

thinned ridge image and sum up all pixels in the row

whose values are one. Then divide the row length by the

above summation to get an inter-ridge width. For more

accuracy, such kind of row scan is performed upon

several other rows and column scans are also conducted,

finally all the inter-ridge widths are averaged to get the D.

Together with the minutia marking, all thinned ridges in the fingerprint image are labeled with a unique ID for further operation.

Fig 4.3

Image

Binarization

Minutiae

Extraction

Save Extracted

Data

Minutiae

Matching

If

Yes Matching

% > 30

Accepted

Person

No

Enhanced

Thinning

Minutiae Post-Processing:

False Minutia Removal: The preprocessing stage does not

usually fix the fingerprint image in total. For example, false

ridge breaks due to insufficient amount of ink and ridge cross-

connections due to over inking are not totally eliminated.

Actually all the earlier stages themselves occasionally

introduce some artifacts which later lead to spurious

minutia. These false minutiae will significantly affect the

accuracy of matching if they are simply regarded as

genuine minutiae. So some mechanisms of removing false

minutia are essential to keep the fingerprint verification

system effective.

Fig 4.4 Flow chart of Proposed Algorithm

Enhanced Thinning

Enhanced Thinning algorithm considers only an eight

neighborhood. However, it has difficulty in

preserving the connectivity of a pattern. To deal with this

problem, we use a 3 x 3 mask. The mask shown in Figure-5

(a and b) is used to indicate the variations of eight

neighboring pixels. A connectivity value is the sum of

each weight of eight directions. After the connectivity

value is calculated and specific conditions are applied, it can

be determined whether the object pixel is to be deleted or

preserved. An essential point is defined as one which

includes a connect point and an end point. The connect

point is a point that its removal causes a disconnectivity in

3 x 3 mask. The end point is a point that has only one of

the eight-adjacent points. Proposed algorithm simply

applies the above definitions to maintain the connectivity in

to the entire image to overcome the deficiencies in

previous parallel thinning algorithms. The proposed

algorithm consists of two steps i.e., Rule 1 calculates

the connectivity value for the entire image step by step

and Rule 2 eliminates non-essential pixels step by step

from the entire image. If all pixels are essential, then the pixel

deletion process is terminated. In sequential image

thinning algorithm the deletion or retention of a (black)

pixel p would depend on the configuration of pixels in a

local neighborhood containing p, and the deletion of p in the

nth iteration depends on all the operations that have been

performed so far, i.e., on the result of the (n- l)th iteration as well as on the pixels already processed in the nth iteration. In proposed parallel image thinning algorithm, the deletion of pixels in the nth iteration would depend only on the result that remains after the (n-l)th iteration, therefore, all pixels can be

processed independently in a parallel manner, for

achieving the better thinned images without excessive erosion and with 8-connectivity.

X 1 X x x x

X Pi 1 1 Pi 1

X 1 X x 1 X

Fig 5 (a) Fig 5 (b)

IV. IMPLEMENTATION AND RESULT

Fig 6.1 Main GUI of proposed Approach

Fig 6.2 Histogram Equalized Image

Fig 6.3 FFT Enhanced Image

Fig 6.4 Binarized Image

Fig 6.5 Direction Finded Image Fig 6.6 Case When Enhanced Thinning Required

Fig 6.7 Result After Enhanced Thinning

Image Previous Proposed

Fig 6.6 ROI extracted Image Result Result

Img 1 26.66 49.24

Img 2 28.84 51.82

Img 3 24.56 48.56

Img 4 27.21 50.25

60

50

40

30 Previous

20 Proposed

10

0

Img 1 Img 2 Img 3 Img 4

Fig 6.7 Thinned Image

V. CONCLUSION

The reliability of any automatic fingerprint system strongly

relies on the precision obtained in the minutia extraction

process. A number of factors damage the correct location of

minutia. Among them, poor image quality is the one with

most influence.

The proposed Enhanced thinning based matching algorithm

is capable of finding the correspondences between minutiae

without resorting to exhaustive research.

There is a scope of further improvement in terms of efficiency and accuracy which can be achieved by improving the hardware to capture the image or by improving the image enhancement techniques. So that the input image to the thinning stage could be made better, this could improve the future stages and the final outcome.