3D To 2D: Conversion For Latent Fingerprint

Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: [email protected]


Volume 4, Issue 5(2), September - October 2015
ISSN 2278-6856

3D to 2D: Conversion for Latent Fingerprint


Vaibhav Jain1 , Ajay Kumar Singh2
1

M. Tech. Student in Computer Science and Engineering at Meerut Institute of Engineering and Technology, Meerut, U.P,
India
2

Pofessor in Computer Science and Engineering at Meerut Institute of Engineering and Technology, Meerut, U.P, India

Abstract
To overcome the problem associated with 2D fingerprints such
as skin deformation we use the concept of 3D fingerprint
scanning. Fingerprints are the combination of ridge and
furrows pattern. Fingerprints are taken by putting on the hard
surface of the sensor like prism, silicon, polymer, index card.
To remove the drawback of traditional 2D fingerprint a new
technology has been introduced called 3D fingerprint. Human
finger is just like 3D object. More fingerprint features and
more real fingerprints will be provided if 3D fingerprints
features. This paper explores 3D fingerprints features and
their application for personal identification.

Index Terms: 2D fingerprint, 3D fingerprint, Data


Acquisition, Latent fingerprint.

1.INTRODUCTION
Today fingerprint recognization system is mostly used in
the forensics labs, police and civilization. We used the
fingerprints recognize in many fields such as whether the
person is criminal or not, the person who access the
information is authentic or not. Fingerprints are the right
choice because no two people have same fingerprints even
the twins has different fingerprints [1]. In legacy system
the fingerprint scans require placing and pressing of
fingers against the hard surface and which results in
partial or degraded quality images. We can raise the latent
fingerprints of the last person who has put his finger on a
sensor are sometimes visible on the surface on the sensor.
And it could be raised through simply breathy on the
sensor. So it is vulnerable. 2D fingerprint does not give
such better result [2]. 3D fingerprints have both depth and
texture information of the fingerprint while the 2D has
only texture information. The 3D fingerprint is free from
distortion as compared to the 2D fingerprints because in
these fingerprints it is free from low pressure and
distortion. Three dimensional systems however, are
usually based on more complex and expensive acquisition
setups [3].

algorithm. [4] introduce Directional Weighted Median


Filter (DWMF) to remove the background noise. [5]
introduce Short Time Fourier Transform (STFT). This
technique estimates all the intrinsic properties of the
fingerprints.

3.POOR WAY OF PLACING THE FINGERPRINT


In the 2D fingerprint system we see that there are many
types of errors while placing the fingerprints on the sensor
[6]. In 3D system these errors are eliminated and the
resulting image is good as compared to 2D fingerprint
system which is shown in figure 1.
Why do we get poor images?
a. Too little pressure:
b. Too much pressure
c. Wet or sweaty
d. Swiped at angle
e. Poor Placement

Fig.1 a.

b.

d.

c.

e.

But in the 3D fingerprint we do not face these types of


errors, because it gives the depth and texture information
both as shown in figure 2.

2.PREVIOUS WORK
The term biometrics has taken the very wide area in the
field of security. Recognition of fingerprints is considered
to be the most powerful technique for security and
authentication. For identifying the fingerprints we need
the better quality of images. Many times fingerprints
found at crime place are not good and we are working on
that. First we eliminate the different type of noise present
in it. We eliminate it by different type of proposed

Volume 4, Issue 5(2), September October 2015

Fig. 2 Good Image


Page 91

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)


Web Site: www.ijettcs.org Email: [email protected]
Volume 4, Issue 5(2), September - October 2015
ISSN 2278-6856
3. DATA ACQUISITION
Data Acquisition is the way of measuring physical or
electrical phenomenon like voltage, current, temperate,
pressure etc. It consists of sensor for measuring the data
from these physical devices. It may be used to convert
physical perimeter to electrical signals. Analog signals
may be converted digital signal by well known instrument
called A to D converter. 3D data acquisition and object
reconstruction can be performed by stereo image pair.
Stereo photogrammetery is based on block overlapping.
Software which could do this is known as Vexcel FOTO
G. Fingerprint can be approximated by cylinders [7]. It is
difficult to capture size variability in vertical and
horizontal direction by best fit cylinder model. So we do
not use parametric method to fit the cylinder to 3D space.
Following step may be involved in 3D acquisition. First,
capture p phase shifted sine wave pattern, capturing can
be done by with the help of projector and camera then
calculate the wrapped phase for the targeted area then go
for segmentation of background with the help of mask.
Use of wrapped phase algorithm for 3D depth matrix use
pre_computed calibration, to remove noise use filter to
obtain 2D equivalent fingerprint. One can use filtering
and fusion. And finally evaluate performance matrix from
2D fingerprint achieved in the above steps which is shown
in figure 3. 2D fingerprint may be distorted if a persons
fingerprint is dirty or his finger is twisted during the
process. 3D fingerprint provide more information as
compared to 2D. 3D is more robust then 2D. 3D can also
be used for ear recognition.

Fig. 3 Process of changing to 3D to 2D fingerprint

Fig.4 3D Palm to 2D Palm

Volume 4, Issue 5(2), September October 2015

Note only fingerprint but also palm can be taken for more
security aspects as shown in figure 4 [7].

4.FEATURE OF 2D AND 3D FINGERPRINT


Traditional fingerprint recognition system uses touch pad
sensor on the other hand 3D uses touchless system. 2D
uses touch based device such as optical and solid state
sensor. Touch based techniques suffer from the following
problem: a). Local Region (ROI) [8] b). Dryness of the
skin c). Sweat d). Dirt e). Humidity in the air. These are
lacuna in 2D [9]. More than one 2D image can be formed
in 3D image. Systems which use multiple view techniques
can only obtain samples that have finger volume with
super imposed image which represent ridge pattern.
Structured light technique and photometric stereo are also
used to identify 3D models of valley and ridge. Non
parametric unrolling algorithm preserves the Geodesic or
Euclidean distance between points. We can denote a 3D
fingerprint in set form such as:
F={( xi, yi, zi, fi)} where i= 1.......N
Where (xi, yi, zi) are the co-ordinates of the ith point of the
fingerprint, fi is the intensity and N is the total number of
points.

5.SIMULATION WORK IN MATLAB


We have made software which matches two images as
shown in figure 5 and 6. Images are taken from the hard
disk which is stored in the format of .jpeg [10, 11]. Here
we are taking 2 same images which show that 100%
matching is done after executing the source code which is
made in MATLAB. The core part of the source code is
given below.
function flag = check_condition( path , end_list ,
branch_list )
[len ,ddummy] = size(path);
curr_x = path(len,1);
curr_y = path(len,2);
flag = 0 ;
if is_a_branch_point(curr_x,curr_y,branch_list) == 1
flag = 1;
elseif is_a_end_point(curr_x,curr_y,end_list) == 1
flag =1;
elseif curr_x == 0 & curr_y == 0
flag = 1;
end;

Fig. 5 Two same images are taken for matching

Page 92

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)


Web Site: www.ijettcs.org Email: [email protected]
Volume 4, Issue 5(2), September - October 2015
ISSN 2278-6856

Fig. 6 Result of above matching

6.CONCLUSION
It is quite helpful for us to convert 3D to 2D fingerprint
because in many cases of latent fingerprint 2D database is
there but 3D database is not yet maintained for checking
criminal activities. By using 3D touchless technique we
can produce many 2D images which can assures us that
matching can be done by any angle. 3D is the latest
technology which is yet to be used and improved by
forensics science department for latent fingerprint
database.

Forensics and Security, Volume 5, No. 4, pp. 750760, 2010.


[7]. Bee Yan Hiew, Andrew Beng Jin Teoh and Ooi Shih
Yin, A Secure Digital Camera Based Fingerprint
Verification
System,
Journal
of
Visual
Communication and Image Representation, Volume
21, No. 3, pp. 219-231, 2010.
[8]. Anju Mohan and Prof. Shilpa P. Kodgire, Touchless
Fingerprint
recognition
using
MATLAB,
International Journal of Innovation and Scientific
Research, ISSN 2351-8014, Volume 10 No. 2, pp.
458-465, October 2014.
[9]. Feng Liu and David Zhang, 3D Fingerprint Reconstruction System Using Feature Correspondences
and Prior Estimated Finger Model, Volume 47,
No.1, Publisher Springer, pp.178-193, 2014.
[10]. Saurabh Yadav and Ajay Kumar Singh, A
Biometric Traits Based Authentication System for
Indian Voting System, International Journal of
Computer Application (IJCA), ISSN: 0975-8887, doi:
10.5120/11002-6185, Volume. 65, No.15, pp. 159162, March 2013.
[11]. Vaibhav Jain and Ajay Kumar Singh, An approach
for Minutia Extraction in Latent Fingerprint
Matching, International Journal, ISSN: 2319 1058,
Volume 6, No. 1 , pp. 51-58, October 2015.

ACKNOWLEDGMENT
We are grateful to Chairman, Director, HoD of MIET
College for their full support.

AUTHOR

References

Prof. (Dr.) Ajay Kumar Singh: Born


in 1974 at Dhanbad (Jharkhand). He
had done B.E (Computer Science &
Engg.) from Kumaon Engineering
College, M. Tech (I.T) Allahabad, Ph.
D (Computer Science & Engg.)
Jaypee University of Information
Technology. Work Experience: He had been in different
institution / university like Radha Govind Engineering
College, Meerut, (U.P), Sir Padampat Singhania
University, Bhatewar, Udaipur, Rajasthan, Jaypee
University of Information Technology, Waknaghat, Solan
(H.P), Mody College of Engineering and Technology,
Lakshmangarh, Sikar, Rajasthan, Regional Engineering
College (Now N.I.T.) Kurukshetra (Haryana), Software
Solution Integrated Ltd. (Delhi), Now he is working with
MIET, Meerut, U.P. He has published several papers in
international Journals like PIER, Asia Magazine EFY,
Elsevier. He has more than 17 years of experience.

[1]. Ruggero Donida Labati, Angelo Genovese, Vincenzo


Piuri and Fabio Scotti, Touchless Fingerprint
Biometrics: A Survey on 2D and 3D Technologies,
Journal of Internet Technology, Volume 15, No.3,
2014.
[2]. Romulo Ferrer L. Carneiro and Jessyca Almeida
Bessa, Techniques of Binarization, Thinning and
Feature Extraction Applied to a Fingerprint System,
International Journal of Computer Applications
Volume 103 - No. 10, p. 4, October 2014.
[3]. Massimo Tistarelli, Stan Z. Li and Rama Chellappa,
Handbook of Remote Biometrics for Surveillance
and Security, ISBN 978-1-84882-385-3, DOI
10.1007/978-1-84882-385-3, Publisher Springer,
2009
[4]. Lin Hong, Yifei Wan and Anil Jain, Fingerprint
Image Enhancement: Algorithm and Performance
Evaluation IEEE Transactions on Pattern Analysis
and Machine Intelligence, Volume. 20, pp. 8, August
1998
[5]. Yiqiu Dong and Shufang Xu., A New Directional
Weighted Median Filter for Removal of Random,
Valued Impulse Noise. Volume 5(9), pp. 2-4, April
2001.
[6]. Yongchang Wang, Laurence G. Hassebrook and
Daniel L. Lau, Data Acquisition and Processing of
3-D Fingerprints, IEEE Transactions on Information

Volume 4, Issue 5(2), September October 2015

Vaibhav Jain: M. Tech student of


Meerut Institute of Engineering and
Technology, U.P, India. Born in 1991
at Shamli.
He had done B.Tech
(Computer Science & Engg.) from
Mangalayatan University Aligarh. His
hobby is searching new things.

Page 93

You might also like