Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more ➡
Download
Standard view
Full view
of .
Add note
Save to My Library
Sync to mobile
Look up keyword
Like this
3Activity
×
0 of .
Results for:
No results containing your search query
P. 1
Enhancement and Minutiae Extraction Of Touchless Fingerprint Image Using Gabor And Pyramidal Method

Enhancement and Minutiae Extraction Of Touchless Fingerprint Image Using Gabor And Pyramidal Method

Ratings: (0)|Views: 1,306|Likes:
Published by ijcsis
Touch based sensing techniques generate lot of errors in fingerprint minutiae extraction. The solution for this problem is touchless fingerprint technology. They do not receive any contact between the sensor & finger. Although they reduce the problems of touch based finger prints, other difficulties explore
such as a view difference problem and a limited usable area due to perspective distortion. To solve this problem, proposed method for touchless fingerprint image enhancement and minutiae extraction is introduced. Image enhancement is mostly required preprocessing system for finger based biometric system. Normally the touchless device is having a single camera and two planer mirrors which reflecting side views of a finger. From this we get three images normally frontal, left and right finger. Experimental result shows that the enhanced images increase the biometric accuracy.
Touch based sensing techniques generate lot of errors in fingerprint minutiae extraction. The solution for this problem is touchless fingerprint technology. They do not receive any contact between the sensor & finger. Although they reduce the problems of touch based finger prints, other difficulties explore
such as a view difference problem and a limited usable area due to perspective distortion. To solve this problem, proposed method for touchless fingerprint image enhancement and minutiae extraction is introduced. Image enhancement is mostly required preprocessing system for finger based biometric system. Normally the touchless device is having a single camera and two planer mirrors which reflecting side views of a finger. From this we get three images normally frontal, left and right finger. Experimental result shows that the enhanced images increase the biometric accuracy.

More info:

Published by: ijcsis on Apr 09, 2011
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See More
See less

09/01/2011

pdf

text

original

 
Enhancement and Minutiae Extraction of Touch less Fingerprint Image Using Gaborand Pyramidal Method
A.John Christopher 
 
Associate Professor, Department of Computer Science,S.T. Hindu College, Nagercoil
Abstract - Touch based sensing techniques generate lot of errorsin fingerprint minutiae extraction. The solution for this problemis touchless fingerprint technology. They do not receive anycontact between the sensor & finger. Although they reduce theproblems of touch based finger prints, other difficulties exploresuch as a view difference problem and a limited usable area dueto perspective distortion. To solve this problem, proposedmethod for touchless fingerprint image enhancement andminutiae extraction is introduced. Image enhancement is mostlyrequired preprocessing system for finger based biometricsystem. Normally the touchless device is having a single cameraand two planer mirrors which reflecting side views of a finger.From this we get three images normally frontal, left and rightfinger. Experimental result shows that the enhanced imagesincrease the biometric accuracy.
Index Terms - pyramidal method, Gabor, touchless fingerprint,thinning, normalization, finger enhancement, adaptive histogram.
I
INTRODUCTION
 
A fingerprint is composed of ridges and valleys.Ridges have various kinds of discontinuity such as ridgebifurification, ridge endings, short ridges, islands and ridgecross over’s. Among this discontinuity, ridge bifurificationand ridge ending are commonly used in fingerprintidentification/verification system and are called minutiae[1].For the processing of fingerprint images, two stages are of pivotal importance for the success of biometricreorganization: image enhancement and minutiae extraction.The traditional fingerprint processing technologies areapplied immediately after sensing. But a better thing is anoptional image enhancement in fingerprint images. Inrealistic scenarios though the quality of a fingerprint imagemay suffer from various impairments, caused by scores, cuts,moist or dry skin, sensor noise, blur, wrong handling of sensor, weak ridge and valley pattern of the given fingerprint,etc. The task of the fingerprint enhancement is to counteractthe aforesaid quality impairments and to reconstruct theactual fingerprint pattern as trace to it original as possible. [2]Fingerprints are traditionally captured based on contact of thefinger on paper or a platen. This often results in partial or degraded images due to improper finger placement, skin
Dr.T.Jebarajan,Principal,V.V. College of Engineering., Tisayanvilai
 deformation, slippage, smearing or sensor noise. Some of thetouch based are shown in fig.1. A new generation of touchless live scan devices that generate three variousrepresentation of fingerprint is appearing in the market. Thisnew sensing technology addresses many of the problemsstated above [3]. From wear and tear of surface coating, toovercome these kinds of problems, a touchless fingerprintsensing technology has been proposed that does not requireany contact between a sensor and a finger. Thus, the fingersand ridge information cannot be changed or distorted as itwill be free of skin deformation. Also, it can capturefingerprint images consistently because it is not affected bydifferent skin conditions or latent fingerprints.
Fig. 1: Distorted images acquired from a touch-based sensor.
Recently, several companies and research groups havedeveloped touchless fingerprint sensors and recognitionsystems [4]–[6]. TST Group developed a touchless imagingsensor (BiRD III) which uses a complementary metal– organic–semiconductor (CMOS) camera, and red and greenlight sources to acquire fingerprint images [4]. Song
et al.
[5]proposed a sensing system with a single charged-coupleddevice (CCD) camera and double ring-type blue illuminatorsto capture high contrast images. Also, Mitsubishi ElectricCorporation proposed another touchless approach
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 9, No. 3, March 201152 http://sites.google.com/site/ijcsis/ISSN 1947-5500
 
transmitting the light through the finger [6], acquiringfingerprint patterns under the surface of skin using light witha wavelength of 660 nm. However, such sensing systems [4]– [6] have an inherent problem as they use only a singlecapturing device, such as CMOS or CCD cameras. whencapturing an image using a single camera, the geometricalresolution of the fingerprint image decreases from thefingerprint center towards the side area [7]. Therefore, falsefeatures may be obtained in the side area and it reduces thevalid and useful region for authentication. Moreover, if thereis a view difference between images due to finger rolling, itreduces the common area between fingerprints and degradessystem performance. To solve this problem, 3-D touchlesssensing systems using more than one view have beenexplored [8]–[11]. TBS [8] used five cameras placed arounda finger to capture nail-to-nail fingerprint images andgenerated a 3-D fingerprint image using the
shape-from-silhouette
method. They then unwrapped the 3-D finger image onto a 2-D image by using parametric andnonparametric models to make rolled-equivalent images [9].Fatehpuria
et al.
[10] proposed a 3-D touchless device usingmultiple cameras and structured light illumination (SLI). Thestructured light patterns are projected onto a finger to obtainits 3-D shape information and 2-D unfolded images aregenerated by applying “
Springs algorithm
” and some postprocessing steps. Also, the Hand Shot ID system wasdeveloped to acquire a 3-D shape of a hand with fingers bystitching images from 36 cameras [11]. Although all thesemethods attempted to solve the problems in touch-basedsensors and acquire expanded fingerprint images with lessskin deformation, they did not raise much interest in themarket because of much higher costs compared toconventional touch-based sensors. Considering the aboveobservations, we adopt a new touchless sensing scheme usinga single camera and a set of mirrors. The mirrors work asvirtual cameras, thus enabling the capture of an expandedview of a fingerprint at one time without using multiplecameras. The device consists of a single camera, two planar mirrors, light-emitting diode (LED)-based illuminators, and alens. Two planar mirrors are used to reflect the left and rightside view of a finger. In this paper, we proposed a newmethod to enhance the touchless finger print and to extractthe minutiae data.
II
 
 
SYSTEM
 
DESIGN
 
To overcome the view difference problem and thelimitation of a single view, some touchless fingerprintingsystems capture several different views of a finger by usingmultiple cameras. However, using multiple cameras increasesthe cost and size of a system. Thus, we adopt a new sensingsystem which captures three different views (frontal, right,and left) at one time by using a single camera and two planar mirrors. Figs. 2(a) and (b) show the prototype and schematicview of the device. As shown in Fig. 2, two mirrors areplaced next to the finger and reflect the right and left sideviews of the finger. Then, the frontal view and two mirror-reflected views are captured by a single camerasimultaneously. A mirror-reflected image is regarded as the“flipped” image taken by a virtual camera placed at adifferent direction compared to the real one. Therefore, wecan capture three different views of a fingerprint using onlyone Camera and also avoid the synchronization problemexisting in multiple camera-based systems. In addition, toobtain high-quality fingerprint images, we need to consider several optical components in order to design the device.
Fig. 2: Proposed device.(a)
 
Prototype of the device. (b) Schematic view of the device.
The specifications of the optical components are as follows:
1)
 
Camera and lens
: We use a 1/3-in progressive scantype CCD with 1024 x 768 active pixels, where thepixel size is 4.65 x 4.65 m. This camera offers asufficient frame rate of 29 Hz, thus avoiding imageblurring caused by typical finger motion. Also, weuse simple equations [see (1) and (2)] to design anadequate lens for our system.
qp
=
 
(1)
 
1 1 1
f p q
= +
 
(2)
Where f is the lens focal length, p and q are the lens-to-object and lens-to-image distances, respectively,and M is the optical magnification. Normally, therequired image resolution for touch-based sensors is500 dpi. Therefore, to ensure a 500-dpi spatialresolution in the fingerprint area and to cover threeview fingerprints, the optical magnificationparameter M, the lens to image distance, and field of view (FOV) are determined as 0.1, 170 mm, and 50x 38 mm, respectively. By doing this, we cancapture three view images with 500-dpi resolution atone time. Also, the depth of field (DOF) of the lensranges from -2.6 to +2.6 mm at a given workingdistance and it normally covers the half depth of afinger.
2)
 
Illumination
:
Considering the reflectance of humanskin to various light sources, we used ring-shapedwhite LED illuminators and a band pass filter whichcan transmit green light to enhance the ridge-to-valley contrast. Also, the illuminators are placedperpendicular to the finger to remove the shadowingeffect. Diffusers are used to illuminate a finger uniformly.
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 9, No. 3, March 201153 http://sites.google.com/site/ijcsis/ISSN 1947-5500
 
 Fig. 3: Overall flowchart of the proposed method
3)
 
Mirror
:
Two planar mirrors are positioned next tothe left and right side of the finger and the mirror size is determined to cover the maximum thumbsize. To provide enough overlapping area betweenfrontal- and side-view images, the angles of themirrors are determined 15 empirically. Also, themirrors can be used as pegs to place a user’s finger firmly on the device.
III
 
 
PROPOSED
 
METHOD
 
In this section, we explain the Enhancement method for synthesizing an expanded fingerprint image from frontal- andside-view images. The overall scheme of the method ispresented in Fig. 3 The method is mainly composed of sixstages (foreground separation, normalisation, Gabor filtering,pyramidal method, thinning, minutiae extraction). Inforeground separation we will do the morphologicaloperation, in normalisation we pre-process the image etc.A)
 
Foreground separation
 Using morphological operation we use the erosionfollowed by dilation, this can be done up to requiredtime. Mathematical morphology is a method of processing digital images on the basis of shape. Adiscussion of this topic is beyond the scope of thismanual. A suggested reference is: Haralick,Sternberg, and Zhuang, "Image Analysis UsingMathematical Morphology," IEEE Transactions onPattern Analysis and Machine Intelligence, Vol.PAMI-9, No. 4, July, 1987, pp. 532-550. Much of this discussion is taken from that article. Briefly, theDILATE function returns the dilation of image bythe structuring element Structure. This operator iscommonly known as "fill", "expand", or "grow." Itcan be used to fill "holes" of a size equal to or smaller than the structuring element. Used withbinary images, where each pixel is either 1 or 0,dilation is similar to convolution. Over each pixel of the image, the origin of the structuring element isoverlaid. If the image pixel is nonzero, each pixel of the structuring element is added to the result usingthe "or" operator.
 
Used with greyscale images,which are always converted to byte type, theDILATE function is accomplished by taking themaximum of a set of sums. It can be used toconveniently implement the neighbourhoodmaximum operator with the shape of theneighbourhood given by the structuring element.Used with greyscale images, which are alwaysconverted to byte type, the ERODE function isaccomplished by taking the minimum of a set of differences. It can be used to convenientlyimplement the neighbourhood minimum operator with the shape of the neighbourhood given by thestructuring element.
B)
 
Normalisation
The process of  removing the effects of the sensor noise and gray-level background due to finger pressure differences. The objective of this stage isdecrease the dynamic range with gray scale betweenridges and valleys of the image. Normalizationfactor is calculated according to the mean and thevariance of the image.
 
Each and every pixel in thefingerprint image has to be processed to find themedian value. The average value of all the pixels iscalculated i.e, the median value. By comparing themedian value with the current pixel the replacementcan be performed
.
Normalization facilitates have the subsequentprocessing steps.Let G (i, j) denote the normalized gray-level value atpixel (i, j). The normalized image is defined asfollows:(3)Where,
0
and
0
VAR
denote the desiredmean and variance value, respectively.Most fingerprint images on a live-scan input deviceare usually of poor quality. The fingerprint image issmoothed with an average or median filter.
C)
 
Gabor filtering 
A Gabor filter is a linear filter used in imageprocessing for edge detection. Frequency andorientation representations of Gabor filter are similar to those of human visual system, and it has beenfound to be particularly appropriate for texturerepresentation and discrimination. In the spatialdomain, a 2D Gabor filter is a Gaussian kernelfunction modulated by a sinusoidal plane wave. The
Foreground
 
separation
 
Gabor
 
filtering
 
Normalization
 
Pyramidal
 
method
 
Thinning
 
Minutiae
 
extraction
 
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 9, No. 3, March 201154 http://sites.google.com/site/ijcsis/ISSN 1947-5500

Activity (3)

You've already reviewed this. Edit your review.
1 thousand reads
1 hundred reads
Rohit Kadlag liked this

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->