Document

ISSN-2349-1841(Online)
Volume 1, Issue 3, May 2015
International Journal of Research Development
& Innovation (IJRDI)
Research Paper
Available online at:www.ijrdi.com
Review on: Finger Vein Recognition
Navpreet Kaur, Er. Varinderjit Kaur
Ramgharia Institute of Engineering. & Technology, Phagwara
[email protected]
HOD, M.Tech (C.S.E)
Ramgharia Institute of Engineering. & Technology, Phagwara
[email protected]
Abstract: Finger vein is a unique physiological
biometric for identification of individuals based on
the physical characteristics and parameters of the
vein patterns in the human. This technology is at
present in use or development for a wide range of
applications
which
includes
credit
card
authentication, security in automobile, employee
time and tracking attendance, computer and
network authentication, security at end points and
automated teller machines. The proposed work
simultaneously obtains the finger-vein and lowresolution finger image images and combines these
two techniques using a better score-level
combination strategy. Analyse the formerly
proposed finger-vein identification methods and
produce a new method that depict it superiority
over prior published efforts. In this thesis
developed and analyzed three new score-level
combinations, i.e. Repeated Line Tracking, Gabor
Filter and K-means comparatively assess them with
more favoured score-level fusion methods to as
precise their effectiveness in the proposed system.
Keywords: Finger Vein Recognition, ROI
Extraction, Image Enhancement, Repeated Line
Tracking, Gabor Filter and K-Means
I. INTRODUCTION
Biometric is the innovation of verifying individuals
utilizing human physiological or behavioural
features for example unique finger impression, iris,
face and voice. Because of the way that a hand
contains lots of data and the data is anything but
difficult to recovered, hand based biometrics for
example finger impression and palm print are the
most well known biometric technologies.
Fingerprint is the most developed hand based
biometric technique where it has been utilized as a
part of numerous applications for quite a long time.
All Rights Reserved
Be that as it may, fingerprint based biometric
framework is vulnerable to forgery because the
fingerprints are effectively presented to the others.
Furthermore, the state of the finger’s surface for
example sweat and dryness can keep a clear
fingerprint pattern from being gotten. This can
degrade the framework’s execution. With respect to
finger knuckle print and palm print based biometric
framework, it is anything but difficult to recreate
following the highlights are outer to the human
body. To defeat the restrictions of ebb and flow
hand based biometric frameworks, finger vein
acknowledgment had been investigated. They
demonstrated that every finger has one of a kind
vein designs so it can be utilized as a part of
individual check. Finger vein based biometric
framework has a few advantages when contrasted
and different hands based biometric systems. First
and foremost, the finger vein example is difficult to
imitate since it is an inside highlight. What's more,
the nature of the caught vein example is not
effortlessly affected by skin conditions.
Finger Image Features
The most common representation used in Finger
image identification is the Galton features. A ridge
can be defined as a single curve segment. The
combination of several ridges forms a finger image
pattern. The small features formed by crossing and
ending of ridges are called minutiae. Ridge Ending
& Bifurcation are taken as the distinctive features
of finger image. In this method the location &
angle of the feature are taken to represent the finger
image & used in the matching process. Together
with there, finger image contains two special types
of feature called core & delta points. The core point
is generally used as a reference point for coding
minutiae & defines as the topmost point on the
P a g e | 83
NavpreetKauret.al. International Journal of Research Development & Innovation (IJRDI)
www.ijrdi.com
Volume 1, Issue 3, May 2015, Pg. 83-87
innermost recurring ridge. The core & delta are
also called the singularity points.
Finger Image Recognition
The uniqueness and changelessness of the finger
image are extremely well-know. Archaeological
artifacts demonstrate those finger images were at
that point utilized by the ancient Assyrians and
Chinese as a manifestation of distinguishing proof
of an individual. The main experimental studies on
finger image date from the late sixteen century but
the essentials of advanced finger image
recognizable proof systems were given toward the
end of nineteenth century. The studies of Sir F.
Galton and E. Henry prompted formally
acknowledge finger image as valid indications of
character by law Enforcement agencies. The
initially Automated Finger image Identification
Frameworks (AFIS) were produced in the 1950s by
the F.B.I. (Federal Bureau of Investigation) in
collaboration with the National Bureau of
Standards the Cornell Aeronautical Laboratory and
Rockwell International Corp.
Finger-Vein Image Pre-processing
The acquired finger images are noisy with
rotational and translational varieties coming about
because of unconstrained imaging. In this way, the
obtained images are initially subjected to preprocessing steps that include:
1) Segmentation of ROI
2) Translation and introduction arrangement and
3)
Image
enhancement
to
concentrate
stable/reliable vascular patterns.
Each of the obtained finger-vein images is initially
subjected to Binarization utilizing a fixed threshold
value as 230 to coarsely localize the finger shape in
the images. A few bits of foundation still show up
as joined with the splendid finger locales,
overwhelmingly because of uneven enlightenment.
The detached and approximately joined districts in
the binarized images are dispensed with in two
stages: First, the Sobel edge indicator is connected
to the whole image, and the subsequent edge guide
is subtracted from the binarized image.
Consequently, the segregated blobs (if any) in the
subsequent images are disposed of from the range
thresholding, i.e., the killing number of associated
white pixels being not as much as an edge. The
subsequent paired veil is utilized to portion the ROI
from the first finger-vein image.
ROI Extraction
The first image is caught with the black undesirable
background. Including the background reduced the
exactness of image.
All Rights Reserved
Figure 1: (a) Original finger vein image, 2(b) finger
edges, and 2 (c) cropped
A special algorithm is delivered to focus the finger
vein image from the background. Three major steps
included in this algorithm. Firstly edge detection is
performed to highlight the finger edge points.
There are two major horizontal lines distinguishing
representing the finger edges as shown in Figure
(b). Secondly combines of edge points are
determined from each of the two major horizontal
lines by checking the lines horizontally. The most
proper cropping points are chosen from the sets of
edge points which must fulfill two conditions: (i)
the range of the set of the edge points is between
35% to 65% of the image height, and (ii) the set of
the edge points is the widest pair among all pairs.
At last the image is cropped vertically at the
cropping points and horizontally at 5% from left
border and 15% from right border. For coordinating
purpose, the size of both registered and input
images are safeguarded to be at the same size. The
ROI of input data image is relying upon the ROI of
registered image. From the identified cropping
points of the input data image, the centre of the
cropping points is calculated. The input data image
is edited at the same height of the registered image
origin from the computed editing points centre.
Image Enhancement
The process of improving the quality of
a digitally damage by manipulating the image
with software. It is simple for instance to make an
image lighter or darker or to increase or decrease
difference.
Advanced
image
improvement
programming also supports many filters for altering
images in various ways. Programs specialized for
image enhancement is sometimes called image
editors.
Repeated Line Tracking
The repeated line tracking technique gives a
promising result in finger-vein identification. The
thought is to follow the veins in the image by
chosen directions as indicated by predefined
P a g e | 84
NavpreetKauret.al. International Journal of Research Development & Innovation (IJRDI)
www.ijrdi.com
Volume 1, Issue 3, May 2015, Pg. 83-87
likelihood in the horizontal and vertical orientations
and the beginning seed is randomly selected and
the entire methodology is over and again
accomplished for a specific number of times.
Literature introduces various researches based on
appearance-based and model-based methodologies
for vein recognition. A concise description of those
recent significant researches is introduced below:
Gabor Filter
It is a linear filter utilized for edge detection.
Frequency and orientation representations of Gabor
filters are like those of the human visual framework
and they have been discovered to be especially
proper
for
texture
representation
and
discrimination. In the spatial domain a 2D Gabor
filter is a Gaussian kernel function balanced by a
sinusoidal plane wave. The Gabor filters are selfcomparable and all filters can be created from one
mother wavelet by dilation and rotation.
Ajay Kumar and Yingbo Zhou (2012) proposed
in their paper exhibits another way to deal with
enhance the execution of finger-vein distinguishing
proof frameworks exhibited in the literature. The
proposed framework simultaneously acquires the
finger-vein and low determination fingerprint
images and consolidates these two confirmations
utilizing a novel score-level combination technique.
They examine the previously proposed finger-vein
recognizable proof methodologies and build up
another methodology that represents it superiority
over prior published efforts.
Formula
𝑔 𝑥, 𝑦; 𝜆,\𝑡𝑕𝑒𝑡𝑎, 𝜓, 𝜎, 𝛾
= 𝑒𝑥𝑝 \𝑓𝑟𝑎𝑐 𝑥 ′2
+ 𝛾 2𝑦
′2
2𝜎 2
\𝑒𝑥𝑝 𝑖 2\𝑝𝑖\𝑓𝑟𝑎𝑐 𝑥 ′ 𝜆
+\𝑝𝑠𝑖
K-means Technique for finger vein
Miura proposed a method that is based on
calculating curvatures in cross-sectional profiles of
a vein image. In each profile the location of the kmeans s is found and those maxima and their width
are taken as the center and the width of the veins
respectively. A new method has been developed to
robustly extract the precise details of the veins by
calculating local k-means s in the cross-sectional
profiles of a vein image. In this method the centre
lines of the veins can be extracted consistently
without being affected by the variations in the
width and brightness of the vein. This method
rectifies the problems found in previous methods
by checking the curvature of the image and
focusing only the centre lines of veins. The centre
lines are obtained by observing the positions where
the curvatures of the cross-sectional profile are
locally maximal. This method of finding the kmeans positions is against the variation in width
and brightness of the vein. The positions are
interconnected with each other and finally the vein
pattern is detected.
The algorithm details are described below:
Step 1: Calculation of the curvatures of profiles.
Step 2: Detection of the centers of veins.
Step 3: Assignment of scores to the center
positions.
Step 4: Calculation of all the profiles.
Step 5: Connection of vein centers.
II. PREVIOUS WORK
All Rights Reserved
Jinfeng Yang and Yihua Shi (2013) proposed a
novel plan for venous region improvement and
finger-vein network segmentation. Firstly a
technique aimed at scattering removal and
directional filtering and false vein information
suppression is put forward to effectively upgrade
finger vein images. At that point to attain the highloyalty extraction of finger-vein networks in an
automated way, matting based segmentation
methodology is exhibited considering the varieties
of veins in intensity and width. Broad analyses are
finally conducted to approve the proposed system.
Huafeng Qin, Sheng Liz, Alex C.Kotz and Lan
Qin proposed in their paper a novel quality
evaluation of finger-vein images for quality control
reason. As a matter of first importance, we isolate a
finger vein image into a set of non-overlapping
blocks. Keeping in mind the end goal to recognize
the nearby vein designs, every piece is anticipated
into the Radon space utilizing an average Radon
transforms. A local quality score is evaluated for
every block according to the curvature in the
comparing Radon space based on which a
worldwide quality score of the finger-vein is
figured and surveyed. Experimental results
demonstrate that our methodology can successfully
recognize the low quality finger-vein images which
are likewise useful in enhancing the execution of
finger-vein recognition.
Kejun Wang, Hui Ma, Oluwatoyin P. Popoola
and Jingyu Li proposed in their paper Accurate
extraction of finger vein pattern is a fundamental
step in developing finger vein based biometric
authentication systems. Finger veins have textured
patterns, and the directional map of a finger vein
image represents an intrinsic nature of the image.
The finger vein pattern extraction method utilizing
oriented filtering technology. Our method extends
P a g e | 85
NavpreetKauret.al. International Journal of Research Development & Innovation (IJRDI)
www.ijrdi.com
Volume 1, Issue 3, May 2015, Pg. 83-87
traditional image segmentation methods, by
extracting vein object from the oriented filter
enhanced image. The best recognition result is
above 90 %.
III. METHODOLOGY
Finger vein identification is one of the most wellknown and publicized biometrics. The steps are:
Step1: It develops a representative data set of
finger images collected from resources.
Step2: It develops a code for loading of finger
images from database.
Step3: It develops a code for the Binarization of
Image and then ROI from finger images.
[1] Ajay Kumar & Yingbo Zhou “Human
Identification Using Finger Images” Published by
IEEE in 2012.
[2] Jinfeng Yang & Yihua Shi “Finger Vein Based
Enhancement and Segmentation” Published by
Springer in 2013.
[3] Vanathi G, Nigarihaa R,Uma Maheswari G
&Sjuthi R “Real Time Recognition System Using
Finger Vein” Published by IJAEEE in 2013.
[4] Wenming Yang, Qing Rao, Qingmin Liao
“Personal Identification For Single Sample Using
Finger Vein Location and Direction Coding”
Published by IEEE in 2011.
Step4: After that it develops code for feature
extraction of Finger Image.
[5] Jinfeng Yang and Xu Li “Efficient finger Vein
Localization and recognition” Published by ISSN
in 2010.
Step5: It develops a code for repeated Line
tracking, Gabor Filter and K-means thus obtain
finger vein image from finger image.
[6] Qin Bin Pan Jian-fei Cao Guang-zhong & Du
Ge-guo “The Anti Spoofing Study of Vein
Identification System” Published by IEEE 2009.
Step6: The same steps used above 2-5 are used for
input finger image.
[7] Yang JF, Yang JL, Shi YH “Finger-vein
segmentation based on multi-channel evensymmetric Gabor filters” Published by IEEE
international conference on intelligent computing
and intelligent systems in 2009.
Step 7: It develop a code to match the database
finger vein images with input vein image and show
the average recognition performance by calculating
FAR and GAR of result obtained.
IV. CONCLUSION
There are some numbers of Finger Vein image
techniques have been proposed earlier but they
were not secure enough and can be temporarily
tampered with so the task was not fulfilled. Finger
Vein and image detection using Repeated Line
Tracking or Gabor Filter alone could not provide
better results. Identity authentication Using
Repeated Line Tracking has been proposed
previously but there have been always need for
better Finger Vein recognition Technique and the
existing Identification Using Finger Vein image
recognition algorithm is costlier. And propose an
enhanced Identification algorithm Using Finger
Vein image which is based on Repeated Line
Tracking; Gabor Filter and K-means.
ACKNOWLEDGEMENT
Thanks to my Guide and family member who
always help and guide me during my dissertation.
Special thanks to my father who always support my
innovative thoughts.
REFERENCES
All Rights Reserved
[8]. David Mulyono & Horng Shi Jinn “A Study of
Finger Vein Biometric for Personal Identification”
Published by IEEE in 2008.
[9] Kejun Wang, Hui Ma, Oluwatoyin P. Popoola
and Jingyu Li “Finger Vein Recognition” Published
by ISSN in 2008.
[10] Naoto Miura, Akio Nagasaka, Takafumi
Miyatake
“Extraction of Finger-Vein Patterns
Using Maximum Curvature Points in Image
Profiles Published by IEEE in 2005.
[11] Naoto Miura, Akio Nagasaka, Takafumi
Miyatake “Feature Extraction of Finger-Vein
Patterns Based on Repeated Line Tracking and its
Application to Personal Identification” Published
by Springer in 2004.
[12] C. Yam, M. Nixon, and J. Carter, “Gait
Recognition by Walking and Running: A ModelBased Approach,” Proc. Asia Conf. Computer
Vision, pp. 1-6, 2002.
[13] E. C. Lee and K. R. Park, “Restoration method
of skin scattering blurred vein image for finger vein
recognition,” Electron. Lett., vol.45, no. 21, pp.
1074–1076, Oct. 2009.
P a g e | 86
NavpreetKauret.al. International Journal of Research Development & Innovation (IJRDI)
www.ijrdi.com
Volume 1, Issue 3, May 2015, Pg. 83-87
[14] Z. Zhang, S. Ma, and X. Han, “Multiscale
feature extraction of finger vein patterns based on
curve lets and local interconnection structure neural
network,” in Proc. ICPR, Hong Kong, 2006, pp.
145–148.
[15] N. Miura,A. Nagasaka, and T. Miyatake,
“Feature extraction of finger vein patterns based on
repeated line tracking and its application to
personal identification, “Mach. Vis. Appl., vol. 15,
no. 4, pp. 194–203, Oct. 2004.
[16] Kefeng Li; “Biometric Person Identification
Using Near-infrared Hand-dorsa Vein Images”.
University of Central Lancashire in collaboration
with North China University of Technology, 2013.
[17] Jinfeng Yang and Xu Li; “Efficient Finger
Vein Localization and Recognition”. IEEE
International Conference on Pattern Recognition,
2010.
All Rights Reserved
P a g e | 87