1. INTRODUCTION:
In India an average yearly report shows
that there are 618176.2 migrants whose intension is not to settle but a
temporary movement of people for the purpose of travel, tourism, pilgrimages.
Such migration gains access for the terrorist to enter the land and cause
damages to life and property. Automation is one of the best approaches
(solution) to increase the border security and safety by using effective
multimodal biometric recognition 4.
The basic idea is that the Iris and Fingerprint
biometric inputs are given by the civilians from air port or bus depot or
harbor or railway stations to the control room. In the control room the
verification process of fusing both the inputs takes place and verifies the
detail about the passengers and then passes on the signal for the migration
verification. Misidentification rate of iris is very less. So here fingerprint
is combined with iris in order to increase the performance of border security
system by means of fusion process. In this system, Passport and unimodal based authentication replaced by
multimodal biometric authentication.
2. EXISTING UNIMODAL APPROACH:
The existing biometric identification
techniques are Unimodal biometric identification and Multimodal biometric
identification with two complete unimodal systems. Unimodal
biometric system consists of its own unique feature extractor and classifier.
Its reliability is decreased because it requires memory footprint, less
accuracy and it have slower processing speed. The block diagram for the
unimodal biometric system is shown in figure 1
Fig.1. Unimodal Biometric System
The block diagram for the existing multimodal
biometric identification system (two complete unimodal) is shown in figure 2.
Fig.2. Existing Multimodal Biometric System
Existing multimodal biometric systems 1 are
always considered to be the combination of two or more complete unimodal
biometric systems. Each unimodal system consists of its own
unique set of feature extractor and matcher thus fusing their scores, require
an additional score normalization setup and a complex fusion approach.The traditional multimodal approach
improves the accuracy and stability of the system over unimodal approach but
this improvement increase the cost of system and larger memory footprint. In
most cases it requires either multiple algorithms or multiple sensors or both
of them.
Drawbacks
in existing methodologies:
Ø It requires larger memory footprint.
Ø High operational cost.
Ø Multiple algorithms are required. Each
unimodal system contains separate matcher (classifier).
Ø Moderate fusion score level.
Ø Slower processing speed.
3. IMPLEMENTATION OF MULTIMODAL BIOMETRIC
IDENTIFICATION 1-8:
The main approach of the proposed system involves
the developing of a fully controlled and automated highly secured border
control using fusion based multimodal biometric identification system without
the availability of two complete unimodal systems. Here using iris and fingerprint as a part of
biometric. Fusing these two input images reduce the chances of hacking. The
basic block diagram for the proposed real time system is shown in figure 3.
Advantages of Proposed System:
Ø Automated
checkout and loss prevention.
Ø Safe journey for
passengers.
Ø High security and
maintenance free.
Proposed Methodology:
One of the major
objectives for the development of the proposed methodology was to demonstrate
that it is possible to design an effective multimodal biometric system without
the availability of two complete unimodal systems. This proposed approach is used to overcome the
problems occurred in existing methodologies and improve the pilot security in
existing system. The block diagram for the proposed methodology is shown in
figure.3. Development of a fingerprint
and iris based multimodal biometric identification system with high score level
fusion that utilizes a weighted Euclidean distance matcher or single hamming
distance matcher 11.
4. METHODOLOGY:
First fingerprint
biometric input is acquired and passed to fingerprint feature extractor. The
processed reference is compared with the templates in the data base using the
provided matcher. In the meantime, the second input is acquired and forwarded
to the iris feature extractor. In the time that the matcher completes the
processing of the first biometric and generate the matching output, the second
biometric input is processed and ready for matching. The same matcher is now
used to compare the iris biometric reference with the templates and generates
the output. The fusion takes place once both matching scores are
available. Proposed fusion based multimodal biometric approach
(figure 3) is for both modalities utilized same classifier is that both output
scores will be same format. The overall proposed system will improves the
security among country borders.
Advantages
of proposed methodology
Ø Eliminating additional normalization
functions.
Ø Improves the processing speed.
Ø Reduces the memory footprint
Ø Simple design process.
Ø Both modalities utilize a single classifier
5.
5. RECOGNITION
PROCESS 11
Proposed
multimodal biometric identification system using the single classifier for both
modalities is that both output scores will be in same format and it also
simplifies the design process. Implementation of multimodal biometric system
consists of the following processes
1. Iris
recognition process.
2. Fingerprint
recognition process.
3. Fusion
process.
The implementation
approach is involved in the
process of acquiring an image of the area containing the text, preprocessing
that image, extracting (segmenting) the individual characters, describing the
characters in a form suitable for computer processing, and recognizing.
Iris
Recognition Process 3:
The Iris
recognition is a very reliable method of personal identification. The iris
image is unique for each person and doesn’t change during the life. The iris
identification system is to automatically recognize the identity of a person
from a new image by comparing it to the human iris patterns denoted with
identity in stored database.
Iris
recognition process composed of four steps.1.First an image containing user’s
eye is captured by CCD or Infrared Camera. (Image Acquisition). 2. Then the
image is preprocessed to normalize the scale and illumination of iris and
localize the iris in the image. (Image Preprocessing).3. Third step, Feature
Extraction representing the iris patterns is extracted. It is done by Using
Daugman’s approach. This Approach including Iris Marking, Cropping, polar
format and Wavelet encoding (converts polar to analog format). This process is
shown in figure 5.
Fig. 5. Iris Feature Extraction process
4. Finally,
the decision is made by means of classification. (Classifier).Classifier refers
to the final decision for recognition or identification person. This isn’t only
matching and comparing iris 5, but also getting the information from iris. In
matching system, it can be tested by using the Weighted Euclidean distance
(WED) algorithm. WED can be used to compare two templates, especially if the
template is composed of integer values. WED gives a measure of how similar a
collection of values are between two templates.
Fingerprint Recognition Process 10:
A fingerprint is comprised of ridges and
valleys. The ridges are the dark area of
the fingerprint and the valleys are the white area that exists between the
ridges. Fingerprint identification
process 10 is having the following steps.
Binarization:
The Binarization
image of fingerprint is shown in figure 6. Binarization is used to convert gray scale image into
binary image by fixing threshold value. The pixel values above and below the
threshold are set to 1 and 0 respectively.
Block Filter:
Block filter
image of fingerprint is shown in figure 7. The Binarization of image is thinned using block
filter to reduce thickness of all ridge lines to a single pixels width to
extract minutiae points effectively. Thinning does not change the location and
orientation of minutiae points compared to original fingerprint which ensures
accurate estimation of minutiae points. Thinning preserves outermost pixels by
placing white pixels at the boundary of the image, as a result first five and
last five rows, first five and last five columns are assigned as value of 1.
Dilation and erosion are used to thin the ridges.
Minutiae Extraction:
Minutiae extraction of fingerprint is shown
in figure 8. The minutiae location and minutiae angles are derived after
minutiae extraction. The termination which lie at the outer boundaries are not
considered as minutiae points, and the crossing number is used to locate the
minutiae points in the fingerprint image. Crossing number is defined as the
half of the sum of difference between intensity values of the adjacent pixels.
If the crossing number is 1, 2, and 3 or greater than 3 then the minutiae
points are classified as termination, normal ridge and bifurcation
respectively.
(a) Original
fingerprint (b) Binarized image
Fig. 6.
Binarization process
Fig. 7. Block Filter process
Fig.8. Minutiae
Extraction process
Fusion Process 6-7:
The fusion of images is the process of combining
two or more (iris & fingerprint) images into single image retaining
important features from each. Recently
multiresolution analysis is one of the acceptable methods to analyze the
remotely sensed images.Several approaches to image
fusion can be distinguished, depending on whether the images are fused in the
spatial domain or they are transformed into another domain, and their
transforms are fused. A new
method based on discrete wavelet transform is proposed here. The DWT based
techniques became popular due to their multi resolution properties. This approach image fusion based on wavelet
decomposition. Here Wavelet transform- Heuristic additive rule based
method is used. That is given below, “I3= ((i1+i2)/2) + ((i1+i2)/100)” It can
show a good position of a function (here this function is the image) in spatial
and frequency space. It is used to display the efficient recognition rate of
combined image. The Iris-Fingerprint
Fusion Process 11 has the following advantages like High Security, Easy to
implement, No training is required, Universally accepted by all and Noise
reduction.
6. RESULT AND DISCUSSION:
Identification
of authorized person based on the classifier output and fusion score level
6-11. In this section we are going to discuss about simulation output for
classifier using MATLAB 9. The details regarding the identification database
and false accept error rates are provided in Matlab coding 9 for all the three
process. In this identification process, one Finger print/Iris image selected
as the verification image input from data base. Then input image is matched
against the entire database images. If the given input matched with the data
base, then the user is identified as an authorized person and the maximum
matching score value will be displayed. If the given input not matched with the
data base then the user is identified as unauthorized person and the minimum
matching score value is displayed with error percentage. The simulation results
of finger print and iris is shown in figure 9&10.
Fig 9.
Simulation matching result for fingerprint
Fig.10.
Simulation matching result for Iris
When
the fusion process score value compared with the individual score values of
both input images, it will improve the efficiency of the average score value.
The simulation result of fusion process is shown in figure 11.
Fig.11.
Simulation result for fusion process
The
simulation result of the proposed fusion based multimodal biometric
identification is clearly shows the improved recognition rate compared with
existing identification method.
7. CONCLUSION:
The proposed
work in this paper has the concept of combining the features of both iris and
fingerprint we can attain very high efficiency and the performance is also
improved. The major advantage of this multimodal biometric identification is
that both modalities utilized the same matcher, low cost with a small memory
footprint and easier for hardware implementation. This recognition can be
implemented in high security areas like airports, War fields, ATM centers etc.,
instead of using passwords. The simulation results clearly show that
the proposed multimodal biometric identification is more secure and efficient
for automation of border security control. So this technique enhances high border security and
thus saves lives and property.
8. REFERENCES:
1.
L. Hong, A. Jain and S. Pankanthi, “Can
Multibiometrics Improve performance?” Proceeding AutoID’99, Summi, NJ, Oct
1999, pp.59-64
2.
A.K. Jain, L. Hong, Y. Kulkarni, “A Multimodal
Biometric System using Fingerprints, Face and Speech”, 2nd
International Conference Audio and Video-based Biometric Person Authentication.
Washington, March 22-24, 1999, pp.182-187
3.
Yunhong Wang, Tieniu Tan, and Anil K. Jain,” Combing
Face and Iris Biometrics for Identity Verification”, 1999, pp.45-54.
4.
A.K. Jain, S. Prabhakar and S. Chen, “Combing
Multiple Matchers for a High Security Fingerprint Verification System”, Vol.
20, No.11-13, 1999, pp1371-1379.
5.
Fabio Roli, Josef Kittler, Giorgio Fumera, Daniele
Muntoni, “An Experimental Comparison of Classifier Fusion Rules for Multimodal
Personal Identity Verification System”, 2002, pp. 76-82.
6.
A. Lumini and L. Nanni, “When Fingerprints Are
Combined with Iris – A case study: FVC2004 and CASIA”, International Journal of
Network Security, vol.4, Jan 2007, pp.27-34.
7.
Karthik Nandakumar, “Multi Biometric Systems: Fusion
Strategies and Template Security”, Ph.D. Thesis, 2008, Michigan State
University,
8.
J.
G. Daugman. High confidence visual recognition of persons by a test of
statistical independence. IEEE Trans. on PAMI, Vol.15 (11), pp. 1148–1161, 1993.
9.
Libor
Masek, Peter Kovesi. MATLAB Source Code for a Biometric Identification System
Based on Iris Patterns. The School of Computer Science and Software
Engineering, the University of Western Australia. 2003.
10.
Z.
Shi and V. Govindaraju, "A chaincode based scheme for fingerprint feature
extraction," Pattern Recogn.Lett., vol. 27, pp. 462-468, 2006
11.
Asim
Baig, Ahmed Bouridane, Faith Kurugollu, and Gang Qu, “Fingerprint Iris Fusion
based Identification System using single Hamming Distance Matcher,” International
journal of Bio-science and Bio-Technology, vol.1, pp.47-57, Dec-2009.
