Professional Documents
Culture Documents
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 1
Webcam Based Fingerprint Authentication for
Personal Identification System
Md. Rajibul Islam, Md. Shohel Sayeed, Andrews Samraj
Abstract: In the networked world there are a huge the image captured by webcam is low resolution image, some
number of systems that need biometric recognition, so at parts of the fingerprint regions are clear but some parts are
present it has become an important issue. For the personal blurred, even it is impossible to extract ridges and valleys.
identification various kinds of vision-based techniques have Third, the lightness of captured fingerprint image is so bright
been proposed earlier. We present a novel one based on and blur. The overall fingerprint identification system using a
visual capturing of fingerprints using a webcam. webcam is composed of preprocessing using gamma
Fingerprint image quality influences deeply the manipulation and gamma correction, image enhancement,
performance of fingerprint identification systems. This feature extraction, and matching algorithm. Fig. 3 shows the
paper presents an improved authentication system using a block diagram of the overall system.
low priced webcam as well as a preprocessing approach
using gamma manipulation and gamma correction This paper presents an overview of the whole interface and a
technique to adjust lightness and intensities of the novel approach to capture fingerprint images using webcam
fingerprint image due to enhance fingerprint image quality. and preprocessing these images in order to improve the
We also implement and test our proposed approach using enhancement and extraction system. We have divided this
the FVC2004 database including the webcam database of paper in the following way: In the next part of this section,
1200 fingerprint images which is obtained by proposed briefly presents an overview of some fingerprint scanner. In
approach and compare the EER (Equal Error Rate), FRR section 2, we describe about the webcam dataset, data
(False Rejection Rate) and FAR (False Acceptance Rate) of collection, overview of the whole system, our contribution
each database. Experiment- al results show that our especially gamma manipulation and gamma correction
approach performs significantly improved and technique of the preprocessing stage before image
comparatively EER, FRR, FAR of the webcam database enhancement and feature extraction in our authentication
are very similar to the FVC2004 database. system. After that, in section 3, we describe the experiments,
discussions using the data obtained from the proposed
Index Terms- fingerprint authentication, web- cam, approach and the data from FVC2004 and a comparative result
fingerprint, fingerprint sensor, gamma manipulation, also presented. Possible future work perspectives is described
gamma correction. in section 4 and by the end of this paper, we present
conclusion.
Biometrics proposes an effectual approach to identify subjects We may not realize it, but the ridges in our fingertips have
because it is concerned with the unique, reliable and stable evolved over the years to allow us to grasp and grip objects
personal physiological features. These features can be: iris [8], with our hands. The ridges and valleys of skin are formed
[9], fingerprints [6], [7], palmprints [10], [11], hand geometry based on genetic and environmental factors, thus, fingerprints
[12], [13], faces [14], [15], voice [16], [17], etc. Most of them are said to be unique from individual to individual. Even
are used for Vision based identification. Voice recognition or identical twins do not share the same fingerprints.
signature verification are the most widely known among the
non-vision based ones. Among these, fingerprint identification A fingerprint sensor is an electronic device used to capture a
has been the most widely browbeaten because of stability, digital image of the fingerprint pattern. The captured image is
usability, and low cost. A fingerprint sensor is necessary for called a live scan. This live scan is digitally processed to create
the commercial fingerprint identification system. Unfortunately a biometric template (a collection of extracted features) which
almost all the modern sensor products are not so cheap and is stored and used for matching. Following are the overview of
available in the market. Therefore we’ve used a low priced some of the more commonly used fingerprint sensor
webcam to fabricate our authentication system. But there are technologies [18].
challenging problems when developing fingerprint
identification system using a webcam. First, the contrast There are two basic methods for scanning fingerprints: Optical
between the ridges and the valleys in images obtained with a scanning and capacitance scanning. Besides Ultrasonic sensors
webcam is low. Second, because of the finger is not flat and also have been used to scan fingerprint.
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 2
Optical epidermis and the area of the sensing element are known
Optical fingerprint imaging involves capturing a digital image values. The measured capacitance values are then used to
of the print using visible light. This type of sensor is, in distinguish between fingerprint ridges and valleys.
essence, a specialized digital camera. The top layer of the
sensor, where the finger is placed, is known as the touch Active capacitance
surface. Besides this layer is a light-emitting phosphor layer Active capacitance sensors use a charging cycle to apply a
which illuminates the surface of the finger. The light reflected voltage to the skin before measurement takes place. The
from the finger passes through the phosphor layer to an array application of voltage charges the effective capacitor. The
of solid state pixels (a charge coupled device) which captures a electric field between the finger and sensor follows the pattern
visual image of the fingerprint. A scratched or dirty touch of the ridges in the dermal skin layer. On the discharge cycle,
surface can cause a bad image of the fingerprint. A the voltage across the dermal layer and sensing element is
disadvantage of this type of sensor is the fact that the imaging compared against a reference voltage in order to calculate the
capabilities are affected by the quality of skin on the finger. capacitance. The distance values are then calculated
For instance, a dirty or marked finger is difficult to image mathematically, using the above equations, and used to form an
properly. Also, it is possible for an individual to erode the outer image of the fingerprint. Active capacitance sensors measure
layer of skin on the fingertips to the point where the fingerprint the ridge patterns of the dermal layer like the ultrasonic
is no longer visible. It can also be easily fooled by an image of method. Again, this eliminates the need for clean, undamaged
a fingerprint if not coupled with a "live finger" detector. epidermal skin and a clean sensing surface.
However, unlike capacitive sensors, this sensor technology is
not susceptible to electrostatic discharge damage. Ultrasonic
Ultrasonic sensors make use of the principles of medical
Capacitance ultrasonography in order to create visual images of the
Like optical scanners, capacitive fingerprint scanners generate fingerprint. Unlike optical imaging, ultrasonic sensors use very
an image of the ridges and valleys that make up a fingerprint. high frequency sound waves to penetrate the epidermal layer of
But instead of sensing the print using light, the capacitors use skin. The sound waves are generated using piezoelectric
electrical current.Capacitance sensors utilize the principles transducers and reflected energy is also measured using
associated with capacitance in order to form fingerprint piezoelectric materials. Since the dermal skin layer exhibits the
images. The two equations used in this type of imaging are: same characteristic pattern of the fingerprint, the reflected
Q wave measurements can be used to form an image of the
C= ………(1) fingerprint. This eliminates the need for clean, undamaged
V epidermal skin and a clean sensing surface.
A
C =∈o ∈r ………..(2) Webcam
d
Webcams typically include a lens, an image sensor, and some
Where support electronics [19]. Various lenses are available, the most
C is the capacitance in farads common being a plastic lens that can be screwed in and out to
Q is the charge in coulombs set the camera's focus. Image sensors can be CMOS or CCD,
V is the potential in volts the former being dominant for low-cost cameras, but CCD
İ0 is the permittivity of free space, measured in farad per meter cameras do not necessarily outperform CMOS-based cameras
İr is the dielectric constant of the insulator used in the low cost price range. Consumer webcams usually offer a
A is the area of each plane electrode, measured in square resolution in the VGA region, at a rate of around 25 frames per
meters second. The higher resolution of 1.3 Megapixel is also
available in the market.
d is the separation between the electrodes, measured in meters.
In this method of imaging, the sensor array pixels each act as The camera pictured to the right, for example, uses a Sonix
one plate of a parallel-plate capacitor, the dermal layer (which SN9C101 to transmit its image over USB. Some cameras -
is electrically conductive) acts as the other plate, and the non- such as mobile phone cameras - use a CMOS sensor with
conductive epidermal layer acts as a dielectric. supporting electronics 'on die', i.e. the sensor built on a single
silicon chip, to save space and manufacturing costs.
Passive capacitance
A passive capacitance sensor uses the principle outlined above
to form an image of the fingerprint patterns on the dermal layer
of skin. Each sensor pixel is used to measure the capacitance at
that point of the array. The capacitance varies between the
ridges and valleys of the fingerprint due to the fact that the
volume between the dermal layer and sensing element in
valleys contains an air gap. The dielectric constant of the
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 3
II. THE WEBCAM DATASET
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 4
be removed. However, this method did not meet the block. (b) Block Diagram of the verification system. Like the enrolment
requirements imposed to a thinning algorithm because it still system after successfully satisfied all the blocks until Feature extraction, live
feature data will verify with the feature data stored with the system database.
left some spurious structures that did not permit a single point
inside a line to have only two neighbors, a ridge-end only one,
C. Our Contribution in this Paper
and a bifurcation three. The conformance to the established
criteria was obtained by the creation of a third test to be run The system we proposed is the most remarkable and automatic
once after the former two are passed to test for certain security device. For this system we connect a webcam of low
conditions in matrices of 3x3 pixels that indicate a spurious priced and low resolution as well, to the computer and apply an
structure that shall be properly modified. identification system to examine the fingerprint it sees, and
then compare this fingerprint image against the minutiae of the
The minutiae extraction process, defined in [3], uses matrices fingerprint belonging to the endorsed user to perform the
of 3x3 pixels to search for typical minutiae, that is: ridge matching process. We attempted both, a relatively high
endings and ridge bifurcations. After extraction minutiae the end Creative unit and a low end Logitech QuickCam, while the
extracted data stores to the system database. Finally for the proposed protocol ready to capture fingerprint images by such
matching process, the live extracted data is to be compared USB-connected webcam and perceived no dissimilarity in
with the extracted data stored in the system database. performance. It gives the impression that a particularly high
resolution fingerprint image does not influence the fingerprint
detection process on the way to identify the necessary features
which we have illustrated below in this section. Alternatively,
in accordance with the changing of lighting circumstances we
detected a difference. Our proposed scheme has more
complexity classifying the features of the fingerprint when the
light source is at the back of the finger, than when the lighting
is at the front or to the side. One more difficulty is to detect
the minutiae information accurately from such low resolution
and poor quality fingerprint image which is captured by low
priced webcam (see Fig. 4).
System (a)
Database
Glass
Webcam Feature
extraction
Preprocessing Fingerprint
stage enhancement
(b)
Yes/ No Fig. 4: (a) Fingerprint image capture by webcam without using crystal clear
Matching glass. (b) Fingerprint image after enhancement.
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 5
captured a shadow of light with the fingerprint image. It lightness range gains more contrast, at the expense of the
doesn’t provide better result to the ridges detection of contrast of the darker colors. At the same time the mean
fingerprint image, due to the lightness of the fingerprint images lightness is decreased, i.e. all new colors are darker than the
captured by low resolution webcam which is so bright and blur. original colors. When γ < 1, the opposite occurs (more
Therefore we have used a preprocessing stage in our system to contrast in the darker colors, less contrast in the lighter colors,
perform gamma manipulation and gamma correction to bend and mean lightness increases).
lightness and intensities of the fingerprint image. The results
are exposed in Fig. 7 and Fig. 8.
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 6
degree of soft clipping and α gives the degree of lightness Where q is the original pixel value and p is the pixel intensity
change (0-100%). as it appears on the display. This relationship is illustrated in
Fig. 6.
In the constant hue plane the original color point can travel
* The fingerprint images captured by webcam, especially during
over the path. This path has the property that chroma ( C ) is
the gamma manipulation they are not corrected for the
constant. In this section we discuss possible steps for gamut- nonlinear relationship between pixel value and displayed
limited manipulations. intensity that is typical for a webcam. This nonlinear
(a) C * = constant, the lightness is manipulated while keeping relationship is roughly a power function, i.e.
Displayed _ Intensity = pixel _ value ∧ gamma .
the chroma C * constant. The maximum and minimum values
are the maximum and minimum lightness L* for this particular
C * value.
(b) C * / L* = constant, the lightness is manipulated while
keeping the ratio C * / L* constant. The minimum value is per
definition 0, the maximum value is the lightness value for
which the line C * / L* intersects the gamut boundary.
Fig. 6: approximate curve to show the intensity response over pixel value
(c) Mapping towards black and white, the lightness is
manipulated in such a way that the point in the This is an approximated curve to show how the intensity
chroma/lightness space moves towards black for a lightness response of a fingerprint image captured by webcam is
decrease and towards white for a lightness increase. nonlinear. Bright colours tend to be displayed too bright. This
(d) Mapping away from black and white, the lightness is can be corrected. The process of adjusting the intensities to
manipulated in such a way that the point in the look correct is known as Gamma Correction.
chroma/lightness space moves away from black for a lightness
increase and away from white for a lightness decrease. The The amount of Gamma Correction we shall call G is usually
maximum and minimum lightness are given by the greater than 1. The range of displayable intensities, P, is
intersections of both lines with the gamut boundary. between 0 and 1. The formula is thus:
Paths along which a color point may move within the constant
hue plane, when applying different steps for gamma
pixel = P ∧ (1 / G ) (7)
manipulations. For each step for an original point, the range for
lightness changes is applied, along with the result for Ȗ = 2 and A G value of 1 gives no Gamma Correction. Higher values
Ȗ = 0.5 and we obtained good outcome for Ȗ=0.9. give more correction.
E. Gamma Correction in Fingerprint image Because values of P must be between 0 and 1, it will have to
divide the intensity by the maximum displayable intensity,
Gamma correction is used to reimburse the nonlinear behavior perform the Gamma Correction, and then multiply up again.
of a poor and low resolution fingerprint image. Using a high pixel = (( P / MaxIntensity ) ∧ (1 / G )) ∗ MaxIntensity (8)
quality Digital camera most often images are already encoded
in gamma corrected, and will appear excellent when displayed
on most video monitors but for the fingerprint image captured
by webcam has to encode in gamma corrected form before
using the enhancement algorithm to obtain better result and to
improve the feature extraction system. However, if a
fingerprint image is stored with a linear scaling it becomes
necessary to correct the image. If the value of gamma for the
webcam is known, then the correction process consists of
applying the inverse of equation (6).
γ
§ q ·
p = 255¨ ¸ (6)
© 255 ¹
(a) (b)
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 7
Table1: The information of dataset
B. Experiments Setup
(c) (d)
We posed 2 experiments. For each experiment, we compared
Fig. 7: (a) Fingerprint image captured by webcam, (b) Grayscale conversion, the FAR and FRR of our webcam database with the rest of 4
(c) Fingerprint image after gamma manipulation and inverting the gamma database which are taken from FVC2004 using TSVM. Both
correction, (d) Enhance part of fingerprint-c. the experiments are done by the method of 5-folder cross
validation, but have differences in the size of test sets and
training sets.
A. Datasets
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 8
Pre-processing Image Feature Enrolment
Enhancement Extraction
Yes
Pre-processing Image Feature Matching /No
Enhancement Extraction
Test Image
Fig. 7: Experiment setup structure using our proposed approach.
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 9
different datasets are compared. And it is clear that the Informatics (Informatics 2007), Kuala Lumpur, Malaysia,
accuracy of the webcam (i.e. low resolution datasets) datasets Nov. 27-28, 2007, pp. T1 (85-91).
using our proposed approach outperforms with the other [8] G. Williams, “IRIS recognition technology", IEEE AES
datasets consistently and significantly system Magazine, pp. 23-29, April 1997.
[9] Daugman, “Recognizing Persons by their Iris Patterns”, In
Biometrics: Personal Identification in Networked Society,
IV. CONCLUSION Kluwer, pp.103-121, 1998.
[10] Guiyu Feng, Dewen Hu, Ming Li and Zongtan Zhou,
We formulated a webcam based fingerprint authentication for “Palmprint Recognition Based on Unsupervised Subspace
personal identification system and presented experimental Analysis”, Lecture notes in Computer Science. Springer
results in this paper. Because of the fingerprint images Berlin / Heidelberg, vol. 3610/2005, pp. 675-678, July 27,
acquired with a webcam are low resolution and sometimes very 2005.
poor quality images; a novel approach was used. We called it [11] Qingyun Dai, Ning Bi, Daren Huang, Dvaid Zhang, Feng
fingerprint pre-processing. Our key contribution of this paper Li, “M-band wavelets application to palmprint recognition
is, we established a pre-processing approach in addition to based on texture features”, International Conference on
gamma manipulation and gamma correction to adjust lightness Image Processing, ICIP2004. Vol. 2, pp. 893 – 896, 24-27
and intensities of the fingerprint image. To estimate the feature Oct. 2004.
locations we used TSVM matching algorithm, which will [12] Jie Wu, Zhengding Qiu, "A Hierarchical Palmprint
evaluate the performance of our proposed scheme regarding Identification Method Using Hand Geometry and
error rates. Our future work in progress is the implementation Grayscale Distribution Features", 18th International
of an optimal matching algorithm to test the quality of the Conference on Pattern Recognition (ICPR'06), pp. 409-
minutiae extraction. Of course, future work will include 412, 2006.
porting the whole algorithm to the hardware coprocessor, [13] Francisco Martinez, Carlos Orrite and Elias Herrero,
possibly using a soft-core processor for certain features. Also “Biometric Hand Recognition Using Neural Networks”,
the interface to the software is going to be ported to a dynamic Lecture notes in Computer Science. Springer Berlin /
link library in order to make it accessible from within many Heidelberg, vol.3512/2005, pp. 1164-1171, June 21,
software development environments, such as Visual Basic, 2005.
Visual C/C++, Delphi, etc. [14] J. Zhang, Y. Yan, and M. Lades, "Face Recognition:
Eigenface, Elastic Matching and Neural Nets”,
Proceedings of IEEE, vol. 85, no. 9, pp. 1423-1435, Sept.
REFERENCE 1997.
[15] R. Brunelli, T. Poggio, “Face Recognition: Features versus
[1] P.D. Kovesi. Matlab functions for computer vision and Templates”, IEEE Trans. on PAMI, Vol. 15, No. 10, pp.
image analysis. 1042-1052, Oct. 1993.
http://www.csse.uwa.edu.au/~pk/Research/MatlabFns/inde [16] J.Picone, “Duration in Context Clustering for Speech
x.html. Recognition”, Speech Communication, Vol.9, pp. 119-128,
[2] T. Y. Zhang , C. Y. Suen, “A fast parallel algorithm for 1990.
thinning digital patterns”, Communications of the ACM, [17] Picone, J.W, “Signal modeling techniques in speech
vol.27 n.3, p.236-239, March 1984 recognition”, Proceedings of the IEEE, Vol. 81, pp. 1215
[3] C.Arcelli and G.S.D.Baja, “A Width Independent Fast – 1247, Sep 1993.
Thinning Algorithm,” IEEE Trans. Pattern Analysis [18] http://en.wikipedia.org/wiki/Fingerprint_authentication
Machine Intelligence, vol. 7, no. 4, pp. 463-474, 1984. [19] http://en.wikipedia.org/wiki/Webcam
[4] Jia Jia, Lianhong Cai, “A TSVM-Based Minutiae [20] Lindsay W. MacDonald and M. Ronnier Luo. Colour
Matching Approach for Fingerprint Verification.” Lecture Image Science Exploiting Digital Media. John Wiley &
Notes in Computer Science. Springer Berlin / Heidelberg. Sons, Ltd. The Atrium, Southern Gate, Chichester, West
IWBRS 2005, vol. 3781, pp. 85-94, 2005. Sussex, England.
[5] J. Dijk and P.W. Verbeek, “Lightness Filtering in Color [21] FVC2004 website,
Images with Respect to the Gamut”, CGIV 2006, Proc. http://bias.csr.unibo.it/fvc2004/download.asp
Third European Conference on Colour in Graphics,
Imaging, and Vision (University of Leeds, UK, June 19-
22), 2006, pp. 330-335.
[6] Anil K. Jain and David Maltoni. Handbook of Fingerprint Md. Rajibul Islam
Recognition. Springer-Verlag New York, Inc, 2003. md.rajibul.islam05@mmu.edu.my
[7] Md. Rajibul Islam, Md. Shohel Sayeed, Andrews Samraj,
“Precise Fingerprint Enrolment through Projection
Incorporated Subspaces based on Principal Component Md. Shohel Sayeed
Analysis (PCA), in Proc. 2nd International Conference on shohel.sayeed@mmu.edu.my
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 10
Andrews Samraj
andrews.samraj@mmu.edu.my
College Science in India 1 : 3 December 2007 Webcam Fingerprint Personal Identification Islam, Sayeed & Samraj 11